Hi - I need help sizing stepper motors for my new Sieg super X3, my intention is to convert to cnc using cadmonkey's design for inspiration, most of the parts that I make are in aluminium or engineering plastics so I feel that using motors rated at over 3nm are a bit over the top, your valued comments would be appreciated.
Regards

Paul

Sieg X2, Sieg X3 Super, Sieg C4

my opinion 465oz to 495oz nema23 for the x and y and 600oz for the z nema34

Thanks for the info Ataxy, seems that i was light on the sizes as the conversion to metric would make them around 360Ncm (3.6Nm) for the X & Y axis and 450Ncm (4.5Nm) for the Z.
the conversion i am undertaking is loosely based on cadmonkeys design but my Z axis will be similar to Syils inasmuch as the stepper will have direct connection to the ballscrew shaft, this in my opinion gives a nice compact drive without having external mounting brackets, also when i bought the mill i asked for a set of ballscrews and shafts - these came ready finished similar to the Syil ones (well the Z axis anyway)

Regards

Paul

Hi Paul!

You really shouldn't pick your motors until you know what drivers and PSU you will be using. Everything has to work together. You will want to wire the motors Bipolar Parallel for best results.

Voltage = power and power = speed. Formula for best motor voltage is 32 times the square root of motor inductance. Try to get as close to that as possible.

Many drivers restrict max voltage. In that case, you should try to find a motor with best voltage equal to or slightly above that restriction.

ON my SX3, I used Keling 425s for X and Y. Their best voltage is 85V, so I used G203s and a 72V PSU. These give good performance--250 IPM rapids, But there are better motors available now.

Today, I would use $49 65V Keling KL23H284-35-4B 387s (back in stock by 2/10/09) for the X and Y, and a $89 640 OZ KL34280-45-8A for the Z axis. I would power these with the $129 KL5010 power supply. I would again go with the G203Vs because there is no better driver anywhere.

If there were a 640 oz N34 motor that would fit it's parameters, I would definitely go with the much less expensive G540, which is a Jr set of 203Vs.

http://geckodrive.com/product.aspx?c=3&i=14460

http://kelinginc.net/KL23H284-35-4B.pdf
http://kelinginc.net/KL34H280-45-8A.pdf

http://kelinginc.net/SwitchingPowerSupply.html

When you are buying the N34 motor, remember that only a single stack motor will fit in the space under the column.

CR.

Wow...

Great post.

I'm buying a Grizzly G0463 mill and am planning to convert it.

Been pouring over this forum, Cadmonkey's and Hoss's sites....

Let's say, it has been a bit of information overload.

Cheers, Bob

Never mind.

i went with nema 23 495oz-in on all 3 axis, 48v supply, cncfusion ballscrew conversion kit, and gecko 203v's.. very happy combination. motors dont heat up too much, and thats why i went 48v rather than something higher, i just didnt see the need, and haven't yet.

You really don't need FAST rapids.
Who wants to cut air above 100 ipm (2500mm/min) anyway.
Well written G code won't be full of wild random seeks anyway.
I leave mine at 60 ipm (1500mm/min). I reduced mine so that my brain reaction time could save on crashes.
If it was a router with a big table, then the rapids may be of more use.

this is true, but i figure it this way. a properly written set of code, shouldnt cause crashes, and secondly you should have an estop button handy thats tied into your control software.

not to mention that this mill is acting as a sort of test bed for the router i'm building.. and it will be big, well, bigger than the mill anyway... so with what i've learned with this about speeds, and whatnot, when i get back to my router, i'll have that much more figured out. I'm also planning to use the same computer, and drivers, to run both my mill and the table router, at different times ofcourse, but i dont see the need to duplicate thoes components right now, i'll never need to use them both at the same time.

this is true, but i figure it this way. a properly written set of code, shouldnt cause crashes, and secondly you should have an estop button handy thats tied into your control software.

not to mention that this mill is acting as a sort of test bed for the router i'm building.. and it will be big, well, bigger than the mill anyway... so with what i've learned with this about speeds, and whatnot, when i get back to my router, i'll have that much more figured out. I'm also planning to use the same computer, and drivers, to run both my mill and the table router, at different times ofcourse, but i dont see the need to duplicate thoes components right now, i'll never need to use them both at the same time.

"you should have an estop button handy thats tied into your control software" - Actually, that's very bad practice. The E-Stop should act as directly as possible on the motor drives themselves, and NOT be dependant on software. It's a good idea to *also* feed an E-stop signal to the software to notify it that the drives have been disabled, but you do not want to be dependant on software (which may have crashed and be the reason you NEED to E-stop) to be required to actually stop the machine.

Regards,
Ray L.

I have wired the E-Stop to the enable line on the stepper drivers.
If the spindle stops dead due to tripping over-current, it needs to issue and instant e-stop so that the feed stops when the cutter is not moving.
This requires a low-speed comparator from the hall sensors on the spindle motor.
Using software to control this safety issue is NOT AN OPTION that should be considered.
The best thing is a safety relay that drops out. That's how the big end of town do it, and then some still do not sense spindle stall.
A cutter not turning, with feed will shatter, and the fragments can become unguided missiles. You won't see it coming!

both are valid points, and i agree, but an estop, weather tied to software or directly to the drives, is still better than no estop. its better than fumbling with a mouse in a panic trying to stop the thing... (yes i made that mistake and learned a lesson)

my one saving grace in that situation was that the way i built my driver box, and where it sits in relation to where i am when i'm running the machine, makes it very easy to kill all the power.

O.K.....to get this thread back on track ;)

I was doing some calculations today about the thrust output for different sizes of stepper motors. I took 25% to be the working output of the motor, and assumed the ballscrew to be 90% efficient. After calculating these values, I compared them by working backwards from the published torque value given by NOOK's website....my values agreed with theirs to within 1.22%

NOTE: These values are for a Ballscrew with 5tpi

200 in-oz = 90lb thrust

300 in-oz = 130lb thrust

400 in-oz = 175lb thrust

500 in-oz = 220lb thrust

600 in-oz = 265lb thrust

700 in-oz = 310lb thrust

So....I guess it depends on how much your saddle/table/workpiece weighs, and how much thrust you need.

Seems like people recommend about 400 in-oz for x-y (175lb) and 600 in-oz (265lb) for z.

If you make it too powerful you will need mechanical end stops, and sophisticated limit switches. The recommended size motors are more than adequate. It will be strong enough to bust a cutter. You invariably limit the feed to stop bending the cutter too much. What you need is the highest spindle speed you can get. That is what limits the performance.

What seems to confuse newcomers like myself is the sheer volume of information that is available both here and on the web, seems that most guys here advocate steppers in the region of 450 - 500 oz for X & Y and 600oz for Z axis.
Now most of my work is on 6030 Aluminium and Delrin with lowish feed rates so I am not wanting high power, also like a lot of folks cash is tight especialy as I have just shelled out for the super X3 so there is a need to keep costs down.
it appears that some UK folks have used steppers sized around 200 - 250oz (1.4 - 1.7Nm) with good results, any ideas on this ?

Using steppers properly sized for best voltage does more than just give high rapid speeds. High rapid speeds are an indicator of motor efficiency. Motor efficiency ensures against lost steps. You can always set mach3 to run the motors slower. Mach3 cannot make inefficient motors run faster without losing steps.

CR.

IMO, These drivers, motors and 48 - 50V PSU will be the best performers on an X3:

http://geckodrive.com/product.aspx?c=3&i=14460

http://kelinginc.net/KL23H284-35-4B.pdf

http://kelinginc.net/KL34H280-45-8A.pdf

http://kelinginc.net/SwitchingPowerSupply.html

CR.

CR,

I noticed in another post that those were on sale for $123 a bit ago....do they do that often? Is there any cheaper place to get them?

Thanks.

-Caleb105-

CR,

I noticed in another post that those were on sale for $123 a bit ago....do they do that often? Is there any cheaper place to get them?

Thanks.

-Caleb105-

Keling sells them for $139 each with motor purchase.

http://kelinginc.net/GeckorDriver.html

CR.

Caleb105.
These are the same motors as were supplied on my factory SX3. They work well.
KL-600-48 48v/12.5A power supply leaves adequate reserve for driving a 4th axis at later date. Make sure that the supply suits the drivers you select.

Hi,

Have a look here for what you need and if you need more help let me know.

http://www.motioncontrolproducts.co.uk

This is where I get all my steppers and drives :) I build my own power supplies.

John

Caleb105.
These are the same motors as were supplied on my factory SX3. They work well.
KL-600-48 48v/12.5A power supply leaves adequate reserve for driving a 4th axis at later date. Make sure that the supply suits the drivers you select.

Neil,

WHICH motors are you referring to?

IMO, These drivers, motors and 48 - 50V PSU will be the best performers on an X3:

http://geckodrive.com/product.aspx?c=3&i=14460

http://kelinginc.net/KL23H284-35-4B.pdf

http://kelinginc.net/KL34H280-45-8A.pdf

http://kelinginc.net/SwitchingPowerSupply.html

CR.


CR,

I've seen you recommend the Keling 387 in-oz (35-4B) before.....why not the 382 in-oz (30-4B)? Is it just the shaft size (1/4 vs. 3/8)?

-Caleb105-

IMO, ... 48 - 50V PSU will be the best performers on an X3:

CR.

Why?

It looks like the G203V can handle 7A and 80VDC....why not a 72V PSU?

-Caleb105-

CR,

I've seen you recommend the Keling 387 in-oz (35-4B) before.....why not the 382 in-oz (30-4B)? Is it just the shaft size (1/4 vs. 3/8)?

-Caleb105-

Shaft is odd, but can be worked around by boring out 1/4 inch couplers. The 382 is a 75V motor, so would require 72V PSU for best performance. The 387 (Back in stock tomorrow?) is a better performer with 48V PSU.

CR.

The 382 is a 75V motor, so would require 72V PSU for best performance. The 387 (Back in stock tomorrow?) is a better performer with 48V PSU.

CR.

CR,

How do you know that it's a 75V motor? Both spec sheets list them as 3.15V, 3.5A.

-Caleb105-

Why?

It looks like the G203V can handle 7A and 80VDC....why not a 72V PSU?

-Caleb105-

$180 for 72V PSU and Keling has no 640 oz N34 motors for 72V--unless you go to the 1200 ounce, which is a little slower.

CR.

CR,

How do you know that it's a 75V motor? Both spec sheets list them as 3.15V, 3.5A.

-Caleb105-

Formula for best/Most efficient motor Voltage is 32 times the square root of inductance in mH. The closer you can get to this, the better.

CR.

Formula for best/Most efficient motor Voltage is 32 times the square root of inductance in mH.

CR.


Yeah....just read that on the G203V manual. My bad.(nuts)


But....the 382 in-oz and the 387 in-oz are both showing 4.1 mH....so how is one rated for a different voltage?

ETA: And if both of these are 4.1mH, then 32 * (sqrt(4.1)) = 65Volts.....isn't 48-50 too low?

-Caleb105-

Yeah....just read that on the G203V manual. My bad.(nuts)


But....the 382 in-oz and the 387 in-oz are both showing 4.1 mH....so how is one rated for a different voltage?

ETA: And if both of these are 4.1mH, then 32 * (sqrt(4.1)) = 65Volts.....isn't 48-50 too low?

-Caleb105-

The 382 is a 5.6 mH motor--hence 75V:

http://kelinginc.net/KL23H2100-30-4B.pdf

There ARE no perfect 48V motors in this power range yet--MOST are much higher Voltage--the idea is to get as close as you can. The 387 is IMO the best available right now.

CR.

The 382 is a 5.6 mH motor--hence 75V:

CR.

CR,

Wow...I just realized that the spec sheets I was switching back and forth between were BOTH for the 387....man....rough day for me.

Thanks for the info!

ETA: Oh, and for the Nema 34, it should be an 8-wire that's wired bipolar parallel right?

-Caleb105-

ETA: Oh, and for the Nema 34, it should be an 8-wire that's wired bipolar parallel right?

-Caleb105-

Eight wire motor can be wired anyway you want. SHOULD be wired BPP for best results. The 387 is a 4 wire motor. You can't change the way it's wired, but internally, it is wired BPP. (4 wire motors are internally wired either BPP or BPS)

The 640 N34 that I recommend:

http://kelinginc.net/KL34H280-45-8A.pdf

IS a 48V motor, so should be very efficient.

CR.

It looks like the Keling 282 in-oz Nema 23 is 2.2mH => 50 volts....but you are also losing about 100 in-oz, or about 45lbs of thrust.

It looks like the Keling 282 in-oz Nema 23 is 2.2mH => 50 volts....but you are also losing about 100 in-oz, or about 45lbs of thrust.

Yep! That is a good range for the lighter X2 table. However, the CNC X2 is more feasable with the much less expensive G540. G540 cannot handle the 4.2A of the 282. (Running 282 on 3.5A G540 would lower available torque to 235 ounces.)

A better choice for X2 and G540 is the 270 ounce 60V KL23H276-28-4B.

CR.

I know it would be a bit more expensive, but would it be better to get the following:

NEMA 23: KL23H284-35-4B - 387 in-oz (works out to 65V motor)

NEMA 34: KL34H280-55-4A - 640 in-oz (works out to 65V motor)

PSU: KL- 6515 - 65V, 15A

The PSU would be $189 compared to $129 for the KL-5010, but then you'd be getting the necessary voltage to all motors.

ETA: It would also give you 15A as compared to the KL-5010's 10A

Thoughts?

-Caleb105-

I know it would be a bit more expensive, but would it be better to get the following:

NEMA 23: KL23H284-35-4B - 387 in-oz (works out to 65V motor)

NEMA 34: KL34H280-55-4A - 640 in-oz (works out to 65V motor)

PSU: KL- 6515 - 65V, 15A

The PSU would be $189 compared to $129 for the KL-5010, but then you'd be getting the necessary voltage to all motors.

ETA: It would also give you 15A as compared to the KL-5010's 10A

Thoughts?

-Caleb105-


This would be an IDEAL combination and I would most wholeheartedly endorse it--IF ONLY KELING ACTUALLY HAD THAT N34 IN STOCK! He claims that nobody ever buys it so he doesn't actually stock it.

Hey! Maybe he HAS it in stock now. Check for availability before ordering, and as in ALL Keling orders: Specify NO SUBSTITUTIONS.

If you have the time to wait--It might be worthwhile to special order it.

CR.

Man....the electronics really add up on this stuff don't they?

Power Supply = $190
Two Nema 23's = $100
One Nema 34 = $90
Three G203V's = $420
----------------------
Total $800

and dont forget all the wires, and connectors and all that little extra stuff that'll add up too...

If i remember correctly, i spent considerably less than 190 for the powersupply.. all you really need is a transformer, a bridge rectifyer, and a couple caps..

Yeah....

I'm a Mechanical Engineering student doing this for my Senior Design project. I should probably go ahead and design my own 65V/15A PSU anyway....would certainly save me a lot of $$$.

Anybody got any links to pre-designed PSU for comparison?

-Caleb105-

This would be an IDEAL combination and I would most wholeheartedly endorse it--IF ONLY KELING ACTUALLY HAD THAT N34 IN STOCK! He claims that nobody ever buys it so he doesn't actually stock it.

Hey! Maybe he HAS it in stock now. Check for availability before ordering, and as in ALL Keling orders: Specify NO SUBSTITUTIONS.

If you have the time to wait--It might be worthwhile to special order it.

CR.

I just sent them an e-mail to confirm wether it was in stock or not.

-Caleb105-

If i remember correctly, i spent considerably less than 190 for the powersupply.. all you really need is a transformer, a bridge rectifyer, and a couple caps..

The transformer is the expensive part. Seems like a Toroidal standard power transformer that outputs 60-66 volts ranges from $130 - $175....makes the already assembled price of $189 not look so bad.

I just sent them an e-mail to confirm wether it was in stock or not.

-Caleb105-


I just ordered one of those N34's that's rated for 65 Volts (KL34H280-55-4A). I then sent him an e-mail stating that I didn't want any substitution (like a 45-4A). He e-mailed me right back saying that he just had some 55-4A's clear customs and now has them in stock. :banana:

ETA: I just did some mock-ups for the Keling 387 in-oz NEMA 23, and the 640 in-oz Nema 34 (55-4A model):

Nema 23:

http://caleb-angie.com/wp-content/uploads/2009/02/nema-23-397oz-motor-pic.JPG

Nema 34:

http://caleb-angie.com/wp-content/uploads/2009/02/nema-34-640oz-motor-pic.JPG

I plan on doing a mockup of the entire X3, as well as all brackets and adapters I make during my project. These are all made in Autodesk Inventor. Let me know if you'd like the files, and I can make them available.

...he just had some 55-4A's clear customs and now has them in stock....

That's EXCELLENT news.

CR.

I found a 65 Volt toroidal transformer for $118

I'm looking into how much it would cost to build my own power supply vs. paying Keling $189 for one of theirs.

I also noticed that Keling Power Supplies are actually just Antek models. ;-)

I found a 65 Volt toroidal transformer for $118

I'm looking into how much it would cost to build my own power supply vs. paying Keling $189 for one of theirs.

I also noticed that Keling Power Supplies are actually just Antek models. ;-)

Isn't 65V awfully high? That'll be over 90V after rectification and filtering - too high for Geckos and most other common motor drivers.

Regards,
Ray L.

Isn't 65V awfully high?

Regards,
Ray L.

The Keling power supply outputs 65V DC

I'm still figuring out exactly what I would need for a transformer and rectifier if I was to build it myself.

ETA: You are assuming the output through a fullwave rectifier would be 1.414 times the value of the input? ....so I would need a transformer that outputs about 46 volts under load? ....maybe 45 volts?

-Caleb105-

The Keling power supply outputs 65V DC

I'm still figuring out exactly what I would need for a transformer and rectifier if I was to build it myself.

ETA: You are assuming the output through a fullwave rectifier would be 1.414 times the value of the input? ....so I would need a transformer that outputs about 46 volts under load? ....maybe 45 volts?

-Caleb105-

Yep! And you probably won't save a great deal by building your own, but it WILL be a learning experience.

Yep! And you probably won't save a great deal by building your own, but it WILL be a learning experience.

Yeah....sigh...I just found a 48V transformer that typically gives out about 63V after passing through a bridge rectifier (according to the data sheet)....only problem is it's $170 just for the transformer.

Oh yeah, and there's always the "Honey I burned down the house!" factor if you screw up.

CR.

Hmmmmm....I found a 63volt / 15amp power supply by Antek for $150.... a little better than $190 by Keling.

Cool! Link please.

CR.

Now I'm really getting into this!

I found a transformer that I'm very interested in. It's got a 45 volt output (which would give me 64 volts after going through a rectifier). It's also got a 18 volt and 12 volt output.

And the best part....it's only $84!!!!

Here's the link to the spec sheet:

http://www.antekinc.com/AN-8445.pdf

All info can be found at www.antekinc.com

I ordered some 50V / 25A full wave bridge rectifiers. I'm going to jump into this and see if I can't make a 65volt / 20A PSU for about $110 all things included! :wee:

I am also going to be doing a Sieg X3 conversion, but I'd like to do it as inexpensively as possible. Here is the combination I'm thinking about:

(Z) Nema 34 - 45V - 640 in oz.

(X,Y) Nema 23 - 45V - 282 in oz.

3 - KL-6050 motor drivers

KL- 5010 500W 50V keling power supply.

Here are the questions I have:

1.) Would it be better to go with 387 in oz. steppers and underpower them? I know 282 in oz. is a little on the low side, but I'd be getting ALL 282 in oz. How much of the 387 in oz would I be getting if they were underpowered?

2.) I notice that the gecko drives are about 140 apiece, whereas the keling ones I have picked out are 60 apiece. Why the big difference?

3.) I have NO clue which BOB to go with. Any advice? Please remember I'm trying to keep this fairly inexpensive.

4.) Any foreseen problems with this combination?

THANKS everyone in advance for the info. This forum is excellent!

I am also going to be doing a Sieg X3 conversion, but I'd like to do it as inexpensively as possible. Here is the combination I'm thinking about:

(Z) Nema 34 - 45V - 640 in oz.

(X,Y) Nema 23 - 45V - 282 in oz.

3 - KL-6050 motor drivers

KL- 5010 500W 50V keling power supply.

Here are the questions I have:

1.) Would it be better to go with 387 in oz. steppers and underpower them? I know 282 in oz. is a little on the low side, but I'd be getting ALL 282 in oz. How much of the 387 in oz would I be getting if they were underpowered?

2.) I notice that the gecko drives are about 140 apiece, whereas the keling ones I have picked out are 60 apiece. Why the big difference?

3.) I have NO clue which BOB to go with. Any advice? Please remember I'm trying to keep this fairly inexpensive.

4.) Any foreseen problems with this combination?

THANKS everyone in advance for the info. This forum is excellent!

I think under-sizing your steppers is a bad idea, given that for only a few dollars more, you can get ones that are guaranteed to be big enough. If they end up underpowered due to low supply voltage, that's a lot cheaper and easier to fix later than having to swap out all the motors. I'd go with more like 495 oz-in. This is a very, very, very common mistake in first conversions - try to save a few $, and end up spending more when you end up doing it all over again. Or, end up frustrated because you create lots of scrap due to random lost steps from running too close to the edge. This is exactly why so many people think steppers are inherently unreliable - they're NOT, when used properly.

On the BOB, the only one I don't think I've ever heard a single complaint about is the PMDX. I've personally had, and continue to have, frustrating problems with my CNC4PC C11 BOB (I'm on my third revision now).

Regards,
Ray L.

Just keep in mind that the 495 in-oz motor really should be ran with 85 volts.....so a 50 volt power supply means you are only giving it about 60% of the voltage it would like.

I think one of the big questions Firestorm has is.....what effect does underpowering a motor have? Do you lose torque? Do you lose max RPM's? Do you lose both?

If you are only giving a 495 in-oz motor 60% of the voltage it wants, are you only getting 297 in-oz (60% x 495) anyway?

Thanks for the input so far on the steppers. Caleb105 nailed it when he revised my question. It seems from reading other forums here that undersizing steppers is a bad idea. However, say I went with the 495 in. oz. steppers and ran them at 50v vs. the 85v they are rated for, what can I expect as far as output? Really Caleb105 said everything my question regarding the motors entails. Also, is 495 in oz. a bit large? It seems most people are using the 387 in. oz. but are still underpowering them.

I understand there is a fine line between wanting to save money, and going too cheap. I really want to end up with a quality CNC, but I can deal with slower feed rates and rapids AS LONG AS the quality of the part is still good. I can always upgrade later if I feel I need faster speeds. My goal is to build the most accurate CNC I can, for the least amount of money.

Thanks again everyone for the input! Sorry if my questions are silly, I am new to this!

Thanks for the input so far on the steppers. Caleb105 nailed it when he revised my question. It seems from reading other forums here that undersizing steppers is a bad idea. However, say I went with the 495 in. oz. steppers and ran them at 50v vs. the 85v they are rated for, what can I expect as far as output? Really Caleb105 said everything my question regarding the motors entails. Also, is 495 in oz. a bit large? It seems most people are using the 387 in. oz. but are still underpowering them.

I understand there is a fine line between wanting to save money, and going too cheap. I really want to end up with a quality CNC, but I can deal with slower feed rates and rapids AS LONG AS the quality of the part is still good. I can always upgrade later if I feel I need faster speeds. My goal is to build the most accurate CNC I can, for the least amount of money.

Thanks again everyone for the input! Sorry if my questions are silly, I am new to this!

Rapids will not be your problem. Your problem will be lost steps when cutting. Fast rapids are easy, and don't require a large motor. When I did my X2 conversion, I started with some really cheap 150 oz-in motors, and even those could do 120 IPM, which is really more than there's any practical use for on a small machine. But cutting, even slowly, resulted in lost steps and scrapped parts.

If you pick too large a motor, and under-power it, all you have to do to fix it is get a different power transformer. If you pick too small a motor, you'll have to replace ALL the motors, and likely then still replace the power supply to get the most out of the new motors. Add in a couple hundred $ in scrapped work and lost time, and an extra $100 now can easily save you 3-5X that down the road. This mistake has been made a thousand times on here, and still people try to get away with it.

Regards,
Ray L.

Well I don't want to make it 1001! Well clearly from what you say, the 282 in oz would be too small. The same goes for the 387 in oz? Does anyone know how to calculate what the actual output of an underpowered motor would be? Also can anyone tell me what the difference between a 60 dollar keling controller and a 139 dollar gecko controller is?

Well I don't want to make it 1001! Well clearly from what you say, the 282 in oz would be too small. The same goes for the 387 in oz? Does anyone know how to calculate what the actual output of an underpowered motor would be? Also can anyone tell me what the difference between a 60 dollar keling controller and a 139 dollar gecko controller is?

At low speed (i.e. - normal cutting rates), the larger motor should still reach very close to peak torque. The high voltage primarily helps at higher speeds, as it enables the drive to saturate the motor coils more quickly, so they spend more of their on-time with peak current flowing. At a non-optimal supply voltage, you'll mostly lose high-speed torque, which should not be a limiting factor on a small machine.

Regards,
Ray L.

theres a lot of hype about max voltages for motors, too much.
its pointless to use an undersize motor because it will rapid faster in theory
you need the torque to make good parts.
the difference between keling and gecco? US labor wages vs chinese labor wages.

EXCELLENT information. Thank you both. So it sounds like if I go with:

(Z) Nema 34 - 45V - 640 in oz.

(X,Y) Nema 23 - 85V - 495 in oz.

3 - KL-6050 motor drivers

KL- 5010 500W 50V keling power supply.

PMDX-122 BOB http://www.pmdx.com/PMDX-122/

That should be a fairly good combo? What do you think of that BOB? Will that BOB work with the Keling controllers?

For an extra $90, you could get three KL-6080 drivers which can handle 80 volts.

ETA: Do you really need 495 in-oz motors (about 225lbs of thrust) for the 23's? I think CadMonkey was the one who used 300 in-oz motors for his 23's and didn't have a problem.

well i'm running the 500 oz-in motors on all 3 axis', running them at 48-50v dc, and i'm not loosing steps, and i can still get plenty fast rapids. infact, i can push on the table with all my might, and i cant make it slow, stop, or drop steps till it hits the hard stop.

I honestly think that your better off in the long run to spend a few more $$ now and get the bigger motors. Its just like what ray was talking about, i've sen it time and time again, someone goes too small, and then they get all mad cause it messes up the parts they try and make.

think about it this way, you can buy a truck with a 4cyl, or an 8cyl engine, for about $100 difference, which one are you gonna buy? some people will get the 4cyl thinking that saving money now is great, but then when they hook thier new boat to it, and it wont pull it out of the driveway, thier gonna wish they had spent the $100.

The 500 oz in motors you are using.. are those Nema 23 or 34? You are using the same motor for the z as the x and y? Where did you purchase them, and what did they run you? I can see from the numerous posts already that I need to go big or go home, but how big is big enough? Clearly, the 282 in oz are too small, but are the 387's big enough? If I use the same motors for all three axes, I can use an appropriate psu for all three. That appeals to me.. if the motors aren't a lot more expensive.

EXCELLENT information. Thank you both. So it sounds like if I go with:

(Z) Nema 34 - 45V - 640 in oz.

(X,Y) Nema 23 - 85V - 495 in oz.

3 - KL-6050 motor drivers

KL- 5010 500W 50V keling power supply.

PMDX-122 BOB http://www.pmdx.com/PMDX-122/

That should be a fairly good combo? What do you think of that BOB? Will that BOB work with the Keling controllers?
i have about the same as that aside from psu and bob, my psu is a 48v so you would be allright with that setup fo a benchtop mill

The 500 oz in motors you are using.. are those Nema 23 or 34? You are using the same motor for the z as the x and y? Where did you purchase them, and what did they run you? I can see from the numerous posts already that I need to go big or go home, but how big is big enough? Clearly, the 282 in oz are too small, but are the 387's big enough? If I use the same motors for all three axes, I can use an appropriate psu for all three. That appeals to me.. if the motors aren't a lot more expensive.

www.Homeshopcnc.com is a great source for stepper motors, and Gecko drives, and they give a discount on the Geckos if you buy the motors at the same time. I used their 495 oz-in NEMA23s and G201s on my X2 conversion, and they were wonderful.

Regards,
Ray L.

i picked up my motors and cap filter board from hubbard. As i remember, i paid about 180 for all 3 motors. ofcourse the first set i got were damaged by UPS, but the seller was awesome, and took care of me.. you can see pics of the broke motors at
http://www.jpcustomcrafts.com/badshipping.htm

currently active auction for the cap board and the seller i got it from:
ebay Item number: 330245166324 (he's had these listed for over a year, and i dont know how many he sells, but if i ever need to build another psu, you can bet i'll consider this first.)

Yes they are all nema 23's each one the same as the last. 4 wire, bipolar, and rated at something like 2.5V and 3.5A if memory serves...
they are of good quality, quiet, and work every time i turn them on.

I just run 1 power supply(hubbard cnc cap board, and got the transformer and the bridge rect. from digikey) and i was done. put it all in a computer case, put digital gauges on it, and called it done. you can see some of my x3 conversion and powersupply/control box here:
http://www.jpcustomcrafts.com/new%20shop/photogalleries/convertingx3tocnc/Album1.htm

if your looking at the digital gauges, the top left is ac line V, bottom left is ac line current, top right is dc V and bottom right is dc current out to the 3 geckos, and yes, this particular computer case is identical to the actual computer i'm useing on the cnc machine(i thought it'd be neat to have them match) and the cases were only $13 from newegg.

just as a side note, i did have to ask michael over at cnc fusion to make my z motor mount special for nema 23, as at the time, he normally only offered nema 34 on the z.

Project,
Nice power supply. I'm planning on going the DIY route myself. I just ordered the toroidal transformer today from Antek....$94 delivered.

Here's a quick deviation for those interested in power supplies.

You know that your wall outlet is "120 Volts"....however, that is 120 volts RMS, but it actually swings from about -170 volts to +170 volts.

Connect this to the right transformer, and it will step it down to about 45 volts RMS, or 64 volts peak-to-peak. Here's a graph I just made of what this looks like: (note that all of this moving happens over the course of about 1/8th of a second)

http://caleb-angie.com/wp-content/uploads/2009/02/45v-rms-ac-voltage.JPG

This AC signal then gets put through a "Full Wave" or "Bridge" rectifier. All the rectifier does is flip the negative part of the AC signal to positive. Here's a graph of both the original input voltage (green), and the new rectified voltage (blue). Note that it still reaches a peak of 64 volts, even though it's 45 volts RMS:

http://caleb-angie.com/wp-content/uploads/2009/02/45v-rms-ac-with-rectified-voltage.JPG

The problem is that the stepper motors want a constant DC voltage. We can "Smooth" out this signal by putting a capacitor in parallel with the motors. I though I would experiment to see what effect different sized capacitors would have on my particular setup. This is what the voltage looks like with a 500uF capacitor. The voltage swings from 62.6 volts down to about 47.8 volts, or about 24% of the max voltage. This is an improvement over the purely rectified voltage, but the stepper motors still wont like this signal:

http://caleb-angie.com/wp-content/uploads/2009/02/smoothed-with-500uf-cap-24-percent-change.JPG

I then kicked the capacitor up to 1mF. That "smoothed" the voltage a bit more. Now it drops from 62.6 volts down to about 54 volts, or about 14% of the maximum voltage. This is better, but you should really have drops of less than 10% for stepper motors:

http://caleb-angie.com/wp-content/uploads/2009/02/smoothed-with-1mfd-cap-14-percent-change.JPG

Finally, I tried a 5mF capacitor. This drastically improved the signal. It now only drops from 62 volts down to about 60 volts, for a total drop of only 3.2%...which makes them stepper motors HAPPY!

http://caleb-angie.com/wp-content/uploads/2009/02/smoothed-with-5mfd-cap-32-percent-change.JPG

Keep in mind that this "Smoothness" was achieved by using only a 5 microFarad capacitor. Project used a 33mF cap, so his signal should be pretty darn straight.

I hope this helps fill some people in on the inner workings of a Power Supply Unit!

Thanks again everyone who has given me such great information. I have a few more questions after doing yet even more (seemingly never ending) research.

I have definitely decided to go bigger with the motors. I can get 3 Nema 34 45V 640 in oz motors for about $240.00. I think I will use the same motor for all three axes. I am also still looking at the 3 - KL-6050 motor drivers.

I have a question now about power supplies. On kelings website, they offer a 48V power supply KL-350-48 for 59.99 and then they offer a 40V power supply at the bottom of the page KL- 3815 for 139.99. What is the difference between these power supplies? I am assuming the more expensive one uses a toroidal transformer while the cheaper one just uses integrated circuits? If not, what IS the difference and why should I spend the extra 80 bucks?!

Also, if my motors are 45V and I use a 48V power supply, would that be acceptable or could I damage them? Would I be better off going with a 40V power supply?

Lastly, if anyone has any more info on BOB's that would be great. I already got a recommendation to use PMDX, but if anyone else has any insight that would be excellent.

firestorm1284.

You made the right decision.
http://www.cnczone.com/forums/showpost.php?p=565954&postcount=62

Why have you gone astray. You like spending money? :drowning:
The more expensive power supplies are switchmode, and almost bullet proof.
Why do you like toroids, heat and using extra space?
The nema34's will look silly on the X and Y axis.
If you use a 3HP spindle motor, stiffen the column, counterbalance the head properly and create a 250KG beast (that has a small table) then you might have an argument for the nema34's. I'm outa here.
There is a case for a more powerful motor on the 4th axis, but a nema34 is still not big enough for that. No stepper is. It needs a servo motor, a drag brake and a shaft resolver, but that is another story and outside the present scope here.

well, you can use the nema 34's on the x and y, but they are physically larger than is needed. It will work, just understand that the larger the rotor diameter, the more likely you limit your top end speed. Bigger diameter motors dont like to go as fast as smaller ones. thats why i staid with nema 23's

something else i wanna get straight. when you say the motors are rated for 45v, is that the proper application voltage? the 20x voltage i mean? or is that what the motors say?

when i was picking my motors, they were rated for something like 2.5V, so with steppers you want your driver voltage to be in the neighborhood of 20x that, so 2.5*20=50V.
what i'm not sure is if the 45v you said is like my 2.5, or if thats the proper app voltage. if its the 2.5 then your gonna want a bunch more voltage on the drivers.

i'm wondering if the specs standard has changed, or something....

also, you can mix motors, so long as thier specs are close enough, so lets say if you wanted to save a little cash, you could get the x and y in a nema23 and then do the z in a 34 if you think you need it. just pay attention to the app voltages and currents, if thier pretty close, then you can run them all off the same powersupply.

just as an example, lets say you got x and y like what i have, and then use another motor for z thats rated at 2.8v, well thats close enough to use 1 powersupply.

just my thoughts on the subject. i would save a little coin and stay with the 23's on x and y. they work.

caleb - dont forget that when you filter a supply like that, as you increase the load current the ripple will come back. so you need to figure out how much ripple is acceptable, at what load current, and then figure the filter caps from there...

2.5 volts DC at the rated DC current
Because we use microstepping drivers, our average current can be about the same DC current and will give the same heating effect, but because of the way these current mode drivers work the drivers require a higher supply voltage to make it all work. The drivers effectively do a DC voltage transformation, (and that is the black magic that makes it all work).
Don't confuse the labeled DC steady state DC figures when applying AC to drive the motor. Just believe the set of rules driver manufacturers have kindly evolved to make our life easy, and follow their application notes in detail. You are welcome to re-invent the wheel. Report back to us your results.

I was estimating the PSU voltage required for each motor by taking 32*Sqrt(inductance), as advised in the Gecko literature.

WOW. Ok, I guess don't go TOO big! That's odd that I got such a reaction to that. The CNC conversion plans I have from fignoggle call for 3 nema 34 1000 in oz motors! That seemed like an incredible amount of overkill though. Let me explain myself though.

The reason why I was going to go with the 3 34's, was because I can get 3 of them for about what it would cost me for 1 34 and 2 23's from a guy on ebay. Then they would all be exactly the same voltage. You are right project, I was multiplying the voltage by 20.

So What I'm hearing is that I should go with a large nema 23 for the x an y and about a 640 Nema 34 for the z.

I guess I also wasn't completely clear on what the difference is between the expensive and inexpensive psu's. I got a "You like spending money" followed by "Why do you like toroids, heat and using extra space?"... not sure what that means although I DO appreciate the post. Maybe you could just clarify that for me please.

And still any help on the BOB would be appreciated.

THANKS again everyone for all the great info.

The drivers TRANSFORM the DC voltage according to speed and everything else. You can mix and match motors and all you will find is that the maximum speed changes. On an X3 the torque will always be able to shear a half inch cutter off with spectacular abandon, regardless of safety glasses. Unless you decide to use it as a planer, or shaping machine, and then it will still work.

The head is over 40KG and NEEDS a big motor.

Toroids etc.. A switchmode supply will be MUCH more efficient 50-100w total losses with drivers I suppose. Linear supply with toroid, double the losses, and you need a very expen$ive filter capacitor, a heat$ink for the rectifier (30 watt$) Lots of highly stre$$ed connections. To $ave how much?

Yes. Find a good BOB. I only know the syil one, so can't help much, but I can give you a schematic of the syil one.

Neil,

It depends on what voltage PS you desire. I am shooting for 65 volts. My options as of right now appear to be either purchase one from Keling (still going to be torroidal) for $189, OR build my own (torroidal) for the following:

Transformer = $94 (shipped)
Rectifier = $5 (shipped)
Capacitor = $5 (shipped)

Add on a heat sink and some connectors and it's built for about $110 - $115....a savings of $75!!!

Although....if you know of a 65V switchmode PSU for less than this, I'd certainly be interested.

No not really green, but a week ago it was 46.5 C (115.7 F) here and I could still use my machine (and did). With a linear supply it would be TOO HOT.
I run my machine 10 hours per day, and the power saving$ add up.
Also the whole shooting match is in the back of the column. No cases, boxes etc. Just a PC.

Over 1000 house, 181 dead, 7000 without a bed. When it was hot last week it was cooler with your hands IN your pockets. I am in the suburbs, but 2 houses a few blocks away burnt.

If you are going to run many hours a day, you might recoup your $75 sooner than you think.

No not really green, but a week ago it was 46.5 C (115.7 F) here and I could still use my machine (and did). With a linear supply it would be TOO HOT.
I run my machine 10 hours per day, and the power saving$ add up.
Also the whole shooting match is in the back of the column. No cases, boxes etc. Just a PC.

Over 1000 house, 181 dead, 7000 without a bed. When it was hot last week it was cooler with your hands IN your pockets. I am in the suburbs, but 2 houses a few blocks away burnt.

If you are going to run many hours a day, you might recoup your $75 sooner than you think.

Neil,

I'm glad to hear that your house was not one of the many that burned to the ground.

As for your comment about recouping the $75....either way they are Torroidal power supplies, using approximately the same amount of power. The only difference is that I can build one myself.

You seemed to be talking about the benefits of Switchmode power supplies. I agree with you that they have benefits, but they also appear to be limited to 48volts. I need 65 volts. Therefore, I have no choice but to go with a torroidal PSU, whether I build it myself or buy it already made.

48v AC * 1.4 will give you 65 volts nicely. Fuse the primary side, not the secondary. The filter cap should never get disconnected from the drivers as they can fail when a fuse blows. All permanent joints on the 65v side.
You should easily get 120 IPM rapids on X and Y, and maybe 85 IPM on the Z.

Thanks for the suggestions Neil. I'll definitely keep those in mind when I build mine.

you dont have to go with a torroid, i went with a brick type transformer(multi-tap) and it only barely gets warm to the touch after a couple hours of running my machine.

as for the heatsink, i'm using the computer case metal body as the heat sink for my rectifyer, and its nice and happy there.

My Keling 72V 20A torroidal PSU barely gets warm to the touch also. I actually think the switching PSUs may get hotter--can THEY run without a fan to cool them?

Gecko recommends non-regulated PSUs as best.

CR.

Crevice....I didn't realize that you were running a 72V PSU. What motors are you powering with that? Did you find a 75V Nema 34?

My Keling 72V 20A torroidal PSU barely gets warm to the touch also. I actually think the switching PSUs may get hotter--can THEY run without a fan to cool them?

Gecko recommends non-regulated PSUs as best.

CR.

Motor controllers are generally happier with unregulated supplies, and they definitely are happier having LOTS of capacitor there. They'll work with switchers, but it's not ideal, as switchers cannot react instantaneously to large changes in current demand as a linear, or unregulated, supply can. There is really no significant advantage whatsoever to using a regulated supply, as compared to a properly-sized unregulated supply - just extra electronics that contributes nothing to the operational characteristics of the machine. As for efficiency, switchers are more efficient, but unless you're running the motors under heavy load a very large percentage of the time (and this is *extremely* atypical), you'll never notice a difference in either heat generated or electric cost. On my knee mill, with three large servos, nothing in the power supply ever even gets warm, much less hot. Whether the transformer is a toroid or a brick makes only a very small difference in efficiency, and zero difference functionally. Go with what you can get at a good price.

Back to the original question - you really should go with the 48V supply, not the 40V. The 40V supply WILL almost certainly reduce your top speed. Make very sure you have enough capacitance on the supply using:

C = 80,000 * I / V

Where I is the total continuous motor current for all motors. So, if using 48V, and three 3.5A motors:

C = 80,000 * (3 * 3.5 * 0.707) / 48 = 12,000 uF

More never hurts, less can. So, go to the next larger size, probably 15,000 or 16,000 uF

When deciding supply voltage, it's not exact. If you end up 5%, or even 10%, over or under the calculated vallue it won't make a huge difference. Just don't get too close to the maximum rating of the driver - allow a little margin for energy dumping on deceleration.

Regards,
Ray L.

Ray, i couldnt agree more. Thats infact why i went with the brick over any other option, it was what had the specs i wanted, at the best price i could find.

the one drawback to having too large of a cap would be the inrush charging current when you first turn it on, if you went big enough it could pop a fuse, or hurt the bridge rectifyer, but it'd have to be a really large cap.

Ray, i couldnt agree more. Thats infact why i went with the brick over any other option, it was what had the specs i wanted, at the best price i could find.

the one drawback to having too large of a cap would be the inrush charging current when you first turn it on, if you went big enough it could pop a fuse, or hurt the bridge rectifyer, but it'd have to be a really large cap.

And that's easily dealt with by having a time-delay relay that connects a current limiting resistor between the rectifier and the cap for a few seconds when power is first applied. After 2-3 seconds, the relay is closed, which bypasses the resistor.

Regards,
Ray L.

yea i thought about that, but really, whos gonna bother to put together enough cap for it to matter? i didnt on mine, and my max inrush current throught the bridge is a very short 31A spike, and considering that my bridge is of the 50A variety, i just dont worrie about it.

besides time delay is more complicated that it needs to be, it'd be easier to just run a simple voltage trigger with a zener, a resistor, and a relay. when the voltage across the cap gets high enough, the relay triggers on and jumps the limiting resistor, Which by the way, depending on its value, may just have to be a fairly large wattage one, either that, or you'll be waiting a long time for the cap to charge..

yea i thought about that, but really, whos gonna bother to put together enough cap for it to matter? i didnt on mine, and my max inrush current throught the bridge is a very short 31A spike, and considering that my bridge is of the 50A variety, i just dont worrie about it.

besides time delay is more complicated that it needs to be, it'd be easier to just run a simple voltage trigger with a zener, a resistor, and a relay. when the voltage across the cap gets high enough, the relay triggers on and jumps the limiting resistor, Which by the way, depending on its value, may just have to be a fairly large wattage one, either that, or you'll be waiting a long time for the cap to charge..

Yeah, but you can buy a self-contained relay with adjustable time-delay built in for a few dollars. Simple is good. The danger isn't so much to the bridge, which can easily be way over-sized, but to the line fuse, switch/relay, etc.. You don't want to have to over-size the line fuse just to accomodate the cap charging.

Regards,
Ray L.

that is true..

do you have any links to where i can find thoes relays? mcmaster? digikey?

that is true..

do you have any links to where i can find thoes relays? mcmaster? digikey?


They're certainly available from DigiKey, Mouser and others, don't know about pricing. But they're not uncommon from surplus places as well, and are often quite inexpensive (under $10).

Regards,
Ray L.

Project,

Where did you get the heat sink that you mounted your G203V's on?

-Caleb105-

Thats actually an old rf load. It had fins on both sides, so i ran one side through the bandsaw, lapped it flat, and voila.

I think I'm getting a little too OCD on my 3D mockups. Here's a model I just made for the G203v:

http://caleb-angie.com/wp-content/uploads/2009/02/g203v-pic.JPG

But where's the little Gecko? You didn't include the little Gecko! :-)

Regards,
Ray L.

Damn you! ;-)

That's the ONE aspect that I couldn't figure out how to get on there.

That's some excellent modeling, caleb! What did you draw that in?

CR.

Autodesk Inventor. Same one I used to make the Stepper motors a few pages back.

But where's the little Gecko? You didn't include the little Gecko! :-)

Regards,
Ray L.

Here you go Ray.

http://caleb-angie.com/wp-content/uploads/2009/02/g203v-pic-with-gecko.JPG

Wow! You da MAN caleb!

CR.

Thanks CR. I know this is somewhat pointless to go to this detail, but it's really forcing me to learn Inventor well.

LBFA is the best learning method.

CR.

Here you go Ray.



That's more like it! :-)

Regards,
Ray L.