Router Forums banner

Power supply for LARGE Nema 23 stepper, New CNC Build

5532 Views 18 Replies 9 Participants Last post by  trikesrcool
Hi all!

Any help is much appreciated in advance.

Unfortunately I con not contribute much to this community yet as I am currently amassing my own knowledge. As I gain knowledge I'll be glad to answer questions from others.

Very new to CNC, but attempting to build my own machine for routing wood to build guitars, solid body electrics. The largest pieces I'd likely ever cut would be less than 20"x25"x2". Most much smaller.

The general set up would be as follows

Likely using Fusion360 or some equivalent to make the G-code

I'll likely use Mach3/4 to deliver the G-code through one of those 4-axis breakout boards that connects via USB, although the machine will be 3 axis

Gantry...Yet to be constructed but I've already purchased rails to build this. Not so worried here, I'll work out a solution.

I already purchased 3 rather large Nema 23 steppers a couple years ago as well. Then we moved and everything got stored.
I now think they are overkill but it's what I have already on hand.

My issue is powering the Larger Nema 23s. Sizing the power supply and drivers.

The motors are the 23HS45-4204S, supposedly the largest nema 23 out there. I think I chose them to make sure the holding torque was adequate to support the weight of the router. Again this was my uneducated self getting order happy a couple years ago and probably going too big.

Specs for the motor
  • Manufacturer Part Number: 23HS45-4204S
  • Motor Type: Bipolar Stepper
  • Step Angle: 1.8 deg
  • Holding Torque: 3.0Nm(425oz.in)
  • Rated Current/phase: 4.2A
  • Voltage: 3.78V
  • Phase Resistance: 0.9ohms
  • Inductance: 3.8mH ± 20%(1KHz)

So 4.2A times 3 is 12.6 amps. 2/3 of that is less than 9 amps

Here is where my question comes in.

Voltage didn't seem as straight forward. There is a voltage listed for the motor of course but it didn't seem this could be added directly to get required voltage.

After some research I found someone using the following formula

V=32√Inductance

After some research..... The origin of this formula was the owner of a company called Geckodrive. Marris Friemannis. The 32 is not arbitrary. It came as a result of much testing, assuring steppers operate under 85°C and do not overheat.

Apply that to my motors, in my case with larger than normal Nema 23s, that's just under 62.4V

So supplying 62.4V at 9A. Voltage X Amps puts me at 550W to 575W

Normally, a 48V dc supply is used with Nema 23s but since mine are larger, they are exceeding this by a little over 25%. Kind of far out of my safety margin.

I think this is a case of "It will work, just do it" and use all the normal things labeled for nema 23 steppers. the 48v supply and the 48v max drives, and 24v 240W supply

Or beef up the power supply and drives to get "Optimal" results and cooler steppers, get a 65v supply capable of 800W max with matching drives.

I'll be routing guitar bodies with this cnc, lots of curves, near constant load on both X and Y axis while cutting, I would think a body could take 30-45 minutes to cut out. So if overheating is possible I think I should avoid it.

My gut tells me to take the "Optimal" approach.

So am I being over cautious? Will the normal nema 23 spec components run what I want to run with no problem?

Thanks again in advance for any advice.



Also at the moment I'm not going to do anything complicated with the "spindle" or router. For now I think I will be okay plugging a normal wood working router in to the wall.... with cable management obviously... and manually engaging the power, then running the G-code. Later I may upgrade to a spindle I can control, or maybe be able to detect something like a broken bit. But for now I'm looking for this machine to only tell the router where to go, only operating the X, Y, and Z axis. With a 5v supply and that breakout board I'm pretty sure I can set up limit switches and a Z axis zero. Beyond those basic features I think I will leave anything else for a later day. I'll just have to ensure I account for not being able to adjust router speed when I decide how much of a bite the router is going to take. I am aware of those limitations but I think for what I need this will be golden.

Thanks again

Also if this is in the wrong category I apologize, this seemed the best fit for my question.
See less See more
1 - 19 of 19 Posts
Welcome to the forum, Dave..
Welcome to the forum Dave.
Hi Dave,

Perhaps I can give you some insight to your situation.

First of all with regard to NEMA motor specs - the most useless value is the voltage. I dont even know why they bother including this in the spec sheets because it is simply the rated current divided by the phase resistance.

The most important value you need to concern yourself with is the rated current. Here is where things can get a bit tricky. Are they talking about peak current or RMS current? With stepper motors coming from China it's hard to tell. I would be conservative and size the drivers for 5 to 6 amps. I always assume they are refering to peak current. The RMS value will always be lower.

Drivers that are rated for 65 volts will be mucho expensive. For your purposes, operating in the 36 to 48 volt range will be more than adequate. Just make sure they can handle the current.

For stepper motor drivers I prefer linear power supplies vs. switching power supplies - but that's just me.

See my thread "Another Build Thread" for how I selected my stepper drivers.

Albert
See less See more
  • Like
Reactions: 1
Normally, a 48V dc supply is used with Nema 23s but since mine are larger, they are exceeding this by a little over 25%. Kind of far out of my safety margin.
I built my CNC and am using 48v power supplies for much larger NEMA 34 stepper motors. As Albert said above that voltage will be just fine.
Thanks all!

I had notifications turned off and did not see these replies, thank you all for the help and advice. Decided to log in and give a summary of what I learned to hopefully help others in the future and found all of your replies.

As it turned out the 65v power supply was out of stock and probably stuck on a boat anyway at the time I was ordering parts so I went with the 48v and matching drivers.

The cost difference between the drivers and the supply was the reason I wanted to inquire about this. If it was worth it I would have gone with the 65v and drivers to match but I did not know if it was worth it or not and had no experience to judge that.

Finished the machine up this week and have done some easy cuts and run some simple programs with it.

All seems fine with the 48v. Running about a 15 minute program the motors get warm but not overly warm. Not even uncomfortable to touch really. If anyone has this question in the future I can say for sure you will be fine with the 48v. Definitely no where near the 85c critical temp or whatever it was I read about. Too much reading and worrying not enough seeing first hand on my part.

I think also you have to consider this is hobby level machine not intended for max production and max speed. If you push the absolute limits of the components maybe you need the 65v.

During set up of the machine I got an idea about what the little nema 23s were capable of by mistake. And man they can really give some speed! Way faster than I would ever use the machine intentionally!

I will likely never push the limits even of the 48v supply.

Super excited to have a working CNC in my shop!

Now..... To figure out how to write cam for a complex component like a guitar neck.... That's a different story haha

Thanks again all!
See less See more
  • Like
Reactions: 2
Now..... To figure out how to write cam for a complex component like a guitar neck.... That's a different story haha
I have yet to do this but I really want to give it a shot at some point. Seems like I've seen some guitar necks generated with F360 so you might find some samples to help with your design and CAD/CAM work.
I have found excellent tutorials on YouTube for modeling a guitar.


This guy walks through the modeling without side tracking at all. It's really straight forward and I was able to duplicate these processes inserting my own measurements and design.

His cam tutorials however I'm finding more difficult to follow. Lack of experience on my part. I will get it, I just need more time to learn the cam side.

Mind you I'm no modeling expert either but I was able to do a passable job following this tutorial.
  • Like
Reactions: 1
Good video so far but I've only watched about 30 minutes. Thanks for sharing this one!
Hi Dave,

Perhaps I can give you some insight to your situation.

First of all with regard to NEMA motor specs - the most useless value is the voltage. I dont even know why they bother including this in the spec sheets because it is simply the rated current divided by the phase resistance.

-snip-

Albert
Albert,

Thank you for emphasizing this important fact.


Rick
I use a 36V 9amp power supply for my 4 motors, Nema23 382oz/in @ 3amps. I worked with an engineer at Automation Technologies to make sure this was appropriate when making my purchases. (Great company by the way, they even called to answer questions I had when I emailed them). The reason this works out is that you won't be in a situation where all 4 motors are pulling their full 3amps at any given time, so the total power draw will be under 9amps.
  • Like
Reactions: 1
I use a 36V 9amp power supply for my 4 motors, Nema23 382oz/in @ 3amps. I worked with an engineer at Automation Technologies to make sure this was appropriate when making my purchases. (Great company by the way, they even called to answer questions I had when I emailed them). The reason this works out is that you won't be in a situation where all 4 motors are pulling their full 3amps at any given time, so the total power draw will be under 9amps.
Mike, I too was recommended using a 36V system for my NEMA23 425on/in @4amp motors. I bet I could have gotten away using 24V PSU with no issues. It would be an excellent experiment.


Rick
Mike, I too was recommended using a 36V system for my NEMA23 425on/in @4amp motors. I bet I could have gotten away using 24V PSU with no issues. It would be an excellent experiment.


Rick
Possibly, as others have said the voltage is the least important number. It really is the amp amount that the steppers are able to draw from that power supply.
Voltage for your stepper drivers is just as important as current since power is the product of voltage times current. In fact, you want to use the highest voltage your drivers can safely take. The reason why stepper motor voltage ratings aren't import is that they are used to define the current based on Ohms Law. With modern stepper drivers, any voltage that your driver can handle is ok.
  • Like
Reactions: 2
Voltage for your stepper drivers is just as important as current since power is the product of voltage times current. In fact, you want to use the highest voltage your drivers can safely take. The reason why stepper motor voltage ratings aren't import is that they are used to define the current based on Ohms Law. With modern stepper drivers, any voltage that your driver can handle is ok.
Phil, I agree to an extent. These H-bridge amplifiers are driven in class D which means the transistors are driven in full saturation using high frequency/high slew rate ON/OFF pulses. When saturated, these transistors have little voltage across them, hence they normally do not overheat even in the most demanding circumstances. Therefore, these amplifiers require little voltage to keep these transistors in saturation. The DC resistance across the motor is only part of the equation, more important is how much inductance the motor has. In layman's terms AC impedance is like resistance, but its value is dependent on frequency and inductance. So, at slow speeds the motor has lower impedance, requiring less voltage to maintain max current. At higher speeds the motor has higher impedance, requiring higher voltage to maintain the max current. Remember these are stepper motors and electrical power requirements are not defined by speed. Since fast speeds are only required for jogging, I think that for CNC router applications, a lower voltage PSU could be applied without consequence. As I mentioned, it's a worthwhile experiment.

Regards,

Rick
See less See more
point well presented. a lower voltage (inadequate) power supply can and will limit the maximum speed of a stepper motor.

so on that note, if when purchasing (one-time) a power supply, would it not be best to purchase the one which will not (likely) become a limiting factor of the stepper motors capabilities?

my history is with servo's, and my cnc has servo's. also, i am not well educated on stepper engineering/values. so just asking...
point well presented. a lower voltage (inadequate) power supply can and will limit the maximum speed of a stepper motor.

so on that note, if when purchasing (one-time) a power supply, would it not be best to purchase the one which will not (likely) become a limiting factor of the stepper motors capabilities?

my history is with servo's, and my cnc has servo's. also, i am not well educated on stepper engineering/values. so just asking...
Until I actually take some measurements, my theory is purely conjecture. The reason I desire to pursue this is that I would like to use a TMC5160 Stepper drivers, but it is rated at [email protected] 3A. At the moment I have a set of No-Name Chinesium Stepper drivers (DM556 clone). The design is inherent to produce lots of unwanted noise that may lead to missteps, not to mention dwarf pulses. It claims it is rated at 5 amp, but I seriously do not think it can run 3.0 amp continuous. The TMC5160 is so silent that you can't hear the steppers even when jogging at full speed. So, when my spindle is creating 89dB of noise, I will be at ease knowing that my stepper motors are only making 40dB of noise. ;-)
Phil, I agree to an extent. These H-bridge amplifiers are driven in class D which means the transistors are driven in full saturation using high frequency/high slew rate ON/OFF pulses. When saturated, these transistors have little voltage across them, hence they normally do not overheat even in the most demanding circumstances. Therefore, these amplifiers require little voltage to keep these transistors in saturation. The DC resistance across the motor is only part of the equation, more important is how much inductance the motor has. In layman's terms AC impedance is like resistance, but its value is dependent on frequency and inductance. So, at slow speeds the motor has lower impedance, requiring less voltage to maintain max current. At higher speeds the motor has higher impedance, requiring higher voltage to maintain the max current. Remember these are stepper motors and electrical power requirements are not defined by speed. Since fast speeds are only required for jogging, I think that for CNC router applications, a lower voltage PSU could be applied without consequence. As I mentioned, it's a worthwhile experiment.

Regards,

Rick
A couple of points, slow speed simply isn't an issue. It is the higher speeds that are. Also, higher speed isn't just for rapids. If your machine is rigid enough to cut at 150 IPM, then you will want to have the power to do so. That means the highest voltage you can safely push through the drivers. And high inductance needs high voltage to get the speed you want. There are limits, though - the 32*sqrt(mH) noted above.

FWIW, rapids are important. Maybe not for a small 3018 style machine but once you get over about 24", jogging speed is important. Plus, rapid speed does effect run time.

And forced air cooling of drivers is super important. While you are right about mosfets in saturation, the transition zone has heating and higher voltages mean longer transition times. a couple of fans is cheap insurance.
  • Like
Reactions: 2
A couple of points, slow speed simply isn't an issue. It is the higher speeds that are. Also, higher speed isn't just for rapids. If your machine is rigid enough to cut at 150 IPM, then you will want to have the power to do so. That means the highest voltage you can safely push through the drivers. And high inductance needs high voltage to get the speed you want. There are limits, though - the 32*sqrt(mH) noted above.

FWIW, rapids are important. Maybe not for a small 3018 style machine but once you get over about 24", jogging speed is important. Plus, rapid speed does effect run time.

And forced air cooling of drivers is super important. While you are right about mosfets in saturation, the transition zone has heating and higher voltages mean longer transition times. a couple of fans is cheap insurance.
Phil,
Yeah, it doesn't hurt to have forced cooling even for a class D amp. I have a 2HP spindle with only a single 420 oz/inch stepper motor for my X axis. The maximum RMS current for the Chinesium DM556 drivers is 4 A, I derate this to 3.5A. Therefore, my performance requirements might not be as demanding as yours. I need to emphasize that a lower supply voltage is not limiting the speed of the motor, but limiting the amount of current at high speed. The exact amount of current loss might be negligible, I don't know yet. It should be fun.

Rick
1 - 19 of 19 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top