Motor drive ratios and speed

[kennected]

I hope I'm in the right place. Wasn't sure since it contains both mechanical engineering and electronics...

I'm rebuilding a large 20 year old CNC router. It has 3 axis (one of which, X, utilizes 2 motors) all of which are nema 34 series steppers.

This machine is solid mechanically but the electronics being as old as they are, will be replaced. I have little factory information and understandably without saying so , they are slow to assist my effort to rebuild one of their "retired" back to life.

I'm considering ClearPath servos, so the question is which model would be best.

Since we're using step/direction, the SDSK series is probably obvious but which stock number? I see the differences in speed and torque between motors in the nema34 choices and see motor torque goes down when speeds go up, so it must be a matter of matching the motor to the mechanical pulley ratios used.
I can add that the current steppers turn 1.75 times per inch of axis travel and the motor pulley turns 4 times to each rotation of the pinion gear. The gantry requires a 30 pound pull (measured with a hanging type scale) to make it move and in this axis' case (x) there are 2 motors.

The CPM SDSK-3421P-RLN (middle range option) would produce over 1000 ipm at its rated speed without giving up all of the torque this model has to offer. With 2 motors on the X axis, would I be better off using a faster rated motor with less torque? I cut plastics, foams, and wood primarily. No aluminum at this point.

Am I on the right track? Thanks, Ken
Options in yellow:







Similar Threads:

• Newbie- 180V DC Motor Speed control Neco speed controller card - PLEASE HELP
• Fanuc AC Spindle Drive speed does not match S-code speed
• Need Help!- pulley ratios : motor to ballscrew
• Simple, 3 Axes with Dual Drive/Motor (per axis), hardware splitter drive Interface.
• Travel Speed Ratios???

[Jim Dawson]

I did a retrofit on a customer router using ClearPath servos, worked out very well. What I did was just match torque and speed rating of the original motors. I was able to get enough information from the data tag on the motor to make a decision.

What router are you rebuilding?

[Al_The_Man]

[COLOR=#000000]I


I'm considering ClearPath servos, so the question is which model would be best.

Since we're using step/direction, the SDSK series is probably obvious but which stock number? I see the differences in speed and torque between motors in the nema34 choices and see motor torque goes down when speeds go up,

Am I on the right track? Thanks, Ken

As far as servo's are concerned they have a much better continuous torque curve compared to steppers, Servo's , like steppers. have maximum torque at zero RPM and generally stays fairly flat over the rated rpm range. Steppers on the other hand do not fare well as rpm increases.
The drive usually has to be built to take higher voltage in order to keep the current (torque) constant.
Servo's generally also have a higher rpm working range.
Another thing to keep in mind when using dual motor for one axis is the ability for each to track correctly, sometimes done with electronic gearing when using servo's.
Al.

[kennected]

I did a retrofit on a customer router using ClearPath servos, worked out very well. What I did was just match torque and speed rating of the original motors. I was able to get enough information from the data tag on the motor to make a decision.

What router are you rebuilding?

Thanks for the replies, gents!

The machine is an AXYZ 6010. Nice sized machine for what we do. Seems solid enough and got it for a song.

I've also ordered UCCNC's UC300ETH and UB1 breakout board. The plan is to gut the box and start from scratch

Do either of you think my ClearPath choice makes any sense? It sounds like, without knowing the engineering, that I can expect an improvement over factory motor and drives without knowing much.

[ger21]

Easiest thing to do is call Clearpath, and tell them what you're doing. They'll tell you exactly what you need.
If it were me, on a big machine, I'd definitely upgrade to the Clearpaths.

[Jim Dawson]

Thanks for the replies, gents!

The machine is an AXYZ 6010. Nice sized machine for what we do. Seems solid enough and got it for a song.

I've also ordered UCCNC's UC300ETH and UB1 breakout board. The plan is to gut the box and start from scratch

Do either of you think my ClearPath choice makes any sense? It sounds like, without knowing the engineering, that I can expect an improvement over factory motor and drives without knowing much.

Nice machine.

I don't know if your ClearPath choices make sense. What are the specs on the original motors? You may gain performance, but without knowing anything about the the original motors, there is no way to answer. There should be a data tag on the motors. Maybe someone else on here knows what the motor are.

Most times torque is more important than RPM. 1000 IPM is great, but if it takes 10 seconds to accelerate to that speed it's useless.

[kennected]

Thanks Ger,.
I gave them all of the above information and here was Abe's reply:

"That’s a real hard question for me to answer. There’s a lot that goes into motor sizing and for a wide range of reasons (some legal), I’m not able to go through all of these calculations. The attached spreadsheet is an excellent reference guide on this topic and I do urge you to go through these web links and even purchase the book – it is a good investment.

For price reasons, I would go with the SDSK series so you clearly are on the right track… Keep in mind that going bigger is not all that much more expensive (from one stack length to the next). But you do really need to go through the calculations.

Sorry I can’t be of more assistance in this domain."

He also sent me several links to videos and tables but I may be slow to understand.
The tag specs are: Allied Motion, P/N 5034-404, DC 1.07V, 3.97A, .27 ohms, 200 SR.
Its from 1996 and according to a helpful tech at AM, they are "custom" so the test data is not readily available but he is trying.

[Al_The_Man]

There is also Graphical sizing programs such as Kollmorgen one.
MOTIONEERING Online | Kollmorgen
Al.

[Jim Dawson]

...........

The tag specs are: Allied Motion, P/N 5034-404, DC 1.07V, 3.97A, .27 ohms, 200 SR.

That is very helpful. Those specs tell me that it is a NEMA 34 stepper motor and the holding torque is in the ~450 oz/in range. From that data, and understanding stepper motors, it is possible to extrapolate a replacement without going through a lot of calculations.

Given that on a good day it may be possible to turn a stepper motor at 1200 RPM, but that is pretty much the maximum, and the torque at that speed is almost nothing. The torque curve on the SDSK is almost flat so you will have near full rated torque throughout the speed range.

If you really need 1000 IPM, then the CPM SDSK-3421P-RLN should work. Personally, I would use the CPM SDSK-3421S-RLN and go for the additional torque and sacrifice some top speed. It will still give you 600 IPM. Either motor will give you a dramatic performance increase over the original setup.

[NIC 77]

I can add that the current steppers turn 1.75 times per inch of axis travel and the motor pulley turns 4 times to each rotation of the pinion gear.

Just trying to understand what you're saying, because this is important to your problem.

So, if the motor turns one complete revolution, the machine travels 1/1.75 = 0.571" Is that correct?

And what is the diameter of the pinion gear? I'm talking about the root diameter, that is to say, the effective diameter relating to # of revs and distance moved, not the overall diameter.

Could it be the distance traveled per rev = 2 x pie x r, so with a 4:1 gear ratio, 0.571 x 4 = 2.284" traveled per rotation of the pinion?

2.284 / 2 x pie = 0.363" pinion radius, which gives a pinion diameter of 0.727" ?

Am I on the right track?

It's anyone's guess at this point. Really, we haven't discussed what accelerations you want, the weights of the moving parts, etc. My guess is not yet, no.

Have you seen the other posts I've done for R&P calcs?

So why are you changing the electronics? Have you even tried the old stuff? Or is it simply, the machine is old, let's start from scratch?

There is also Graphical sizing programs such as Kollmorgen one.
MOTIONEERING Online | Kollmorgen
Al.

I've tried a couple of different programs. They didn't help. The design was automation factory based, and the user interface, a PITA. I have made my own spreadsheets to do calculations, and the math isn't too hard. I have found these more relevant to answering the questions that I want answered in regards to acceleration, weights, and torque vs speed graphs.

Also, I found the programs I tried were simply there to try and sell their products to factories. Not to try and help DIY CNCers with their calcs. correct me if I'm wrong, I don't recall if this is one of the programs I tried.

Perhaps the OP will have a different experience with them?

Have you used this program yourself? If so, how did you find it?

The torque curve on the SDSK is almost flat so you will have near full rated torque throughout the speed range.

SDSK-3421P-RLN

Not Flat

https://www.teknic.com/model-info/CPM-SDSK-2341P-RLN/

SDSK-3421S-RLN

Also not Flat

https://www.teknic.com/model-info/CPM-SDSK-3421S-RLN/

Well, the continuous power is flat, but the peak power looks like a stepper. I never did get a straight answer from Teknic about using peak power for acceleration. The truth is, IMO, it depends on the geometry you're cutting, as to the duty cycle that peak power will be used for accel. Something Teknic can't answer because it's outside of their power to know those specifics.

Another thing the OP might consider is DMM 750 watt servos.

OR

Use the steppers that came with it, they worked well enough before.

I'd gladly trade my Nema 34's for some clearpaths, but let's not kid ourselves, they are a premium product. Lots of $$$. If you want to waste your money for something that isn't needed I will gladly take a donation through paypal.

And, if I was going to upgrade for that kind of $$$, based on the torque vs speed graphs, the DMM 750W on a Nema 34 frame are my current choice. Convince me otherwise.

[Al_The_Man]

Also, I found the programs I tried were simply there to try and sell their products to factories. Not to try and help DIY CNCers with their calcs. correct me if I'm wrong, I don't recall if this is one of the programs I tried.
Have you used this program yourself? If so, how did you find it?
.

I have used one of the original versions of Kollmorgen a while ago now, not the recent editions, but I just used it to get in the ball park and get a feel for the demands due to motor to load inertia ratio etc, Yes most suggest their product as a result, but they usually show the torque requirements etc.
Al.

[hanermo]

This is both easy and hard.
The *right* answer is very easy.

The cheapest servos in your range, smallest, will be vastly better than the current steppers.
A 300 oz-in Nema34 stepper will only "deliver" about 1/5 or 20% of torque at top speed, typically used, 20 years ago. Somewhere near 300-500-700 rpm, mostly.
So it is about a 60 oz-in motor for practical purposes.

Imho.
Buy the cheapest servo from your selection, get fantastic results.

BUT ... !!!
Can the controller output the pulse stream ?
Usually, not.
Ok, the UCxxx can, done.

Do not increase acceleration a lot.
Do not increase top speed a lot.
Do rebuild/test the screw mounts, if budget allows.

Thanks for the replies, gents!

The machine is an AXYZ 6010. Nice sized machine for what we do. Seems solid enough and got it for a song.

I've also ordered UCCNC's UC300ETH and UB1 breakout board. The plan is to gut the box and start from scratch

Do either of you think my ClearPath choice makes any sense? It sounds like, without knowing the engineering, that I can expect an improvement over factory motor and drives without knowing much.

[kennected]

So much great help!

NIC 77- Yes, the math holds up, One rev = right around .571" of travel. With my HF calipers, the pinion gear (16 tooth) size is about .68" diameter, not measuring any of the tooth geometry. The rack pitch is about .156" (tooth center to next tooth center), so we are in the same park.
I like the idea of changing out the electronics to know that I can rely on this machine when I need to. When I manually pull it around, it feels mechanically sound and tight, so new electronics should make it easier for simple me to connect and run without trying to figure out the 2- RJ485 connections and whatever else that is "proprietary". There was no PC and no software with it but that was pretty much what I had with my last build like this. I figure that I can make this thing run and maybe be reliably productive for about $8,000. total. Machine, new electronics, and ClearPath motors. So, though they might be relatively pricey I'm still thinking ClearPath because the curiosity is killing me. I have some other motion projects that could use really predictable linear movement and this will indoctrinate me, I guess.

hanermo- Thanks, I agree.

Also, I see that these ClearPaths are looking for 75 volts. That's a lot to me. Does anyone suggest a power supply that isn't as pricey as those on the ClearPath site?

[NIC 77]

I looked for some pics online, so this AXYZ 6010 is quite large?

Do you have the twin spindles on it?

The entire Z axis assembly, what does that weigh? That's got to be several hundred pounds at least? And for the part on the Z that goes up and down, that's pneumatically counter balanced is it?

Any idea what the gantry weight is? That would include the Z axis assembly, all the motors, all the little bits and pieces that actually move when the gantry moves. Everything that goes back and forth.

I'm guessing it's north of 800 lbs? No idea what size and thicknesses of steel are there. Can you come up with a good estimate for the weights?

Also, do you have pics of the 4:1 belt reduction or any specs on that? Trying to figure out what the inertia is on it. It's usually published for a planetary gear.

I ask these questions because I'm contemplating making some graphs for you like I have done previously for other people. That could help you with making your choices, but I don't yet have enough info to do that.

I'm guessing whatever you buy will need to be on a Nema 34 frame to keep the same mounts.

So, the clearpaths in N34, a servo on a N34 frame like a DMM 750 on N34, or an actual N34 stepper.

[Jim Dawson]

Also, I see that these ClearPaths are looking for 75 volts. That's a lot to me. Does anyone suggest a power supply that isn't as pricey as those on the ClearPath site?

I am running 4 of the ClearPath motors on one of these, working great for my customer.

https://www.automationtechnologiesin...vac-or-230vac/

[ger21]

AC servos like the DMM's have different frame sizes than steppers, and would not be a simple drop in replacement like the Clearpaths.

Here's an alternative power supply:
PS-15N73 - 1500W 73V Power Supply - AnTek Products Corp

With servos, rpm is dictated by the voltage. If you were to run them on 48V, you'd only get roughly 2/3 the rated rpm.

[NIC 77]

AC servos like the DMM's have different frame sizes than steppers, and would not be a simple drop in replacement like the Clearpaths.

They have two models that are advertised as being on a Nema 34 frame. Are you saying that those would not be drop in replacements for a Nema 34?

[1Jumper10]

The IPC500 ClearPath power supply makes use of regenerative "braking". Effectively using the power generated by one motor as its decelerating to power the other motor(s) as they are accelerating. Not sure if the other power supplies do that and more importantly I don't know how much that matters. But I did ask Teknic tech support about the power supply being only 500 watts and if that was enough for my machine. They explained that the regenerative feature helps a lot and on a 3 axis router you would only see maximum power demand if all 3 axis are accelerating, such as in a slicing maneuver.

If you get the CP's, with a 4:1 drive ratio my first choice would be their SDSK3421S if that will get you enough top speed. If not then the 3421P would probably work fine. One thing to remember with the CP's if you are going to use them on a dual driven gantry you'll have to put some effort into tuning them. The auto tune feature wont work right for the gantry. The X and Z will be no problem.


DDgitfiddle just tuned some SDSK3421S servo's on his FineLine Automation build and he had a great post about how he tuned his. See post #59 here -> http://www.cnczone.com/forums/fineli...m-posts-3.html

[Al_The_Man]

They have two models that are advertised as being on a Nema 34 frame. Are you saying that those would not be drop in replacements for a Nema 34?

Keep in mind that NEMA rating is a frame mounting size standard, it has no direct relationship with torque etc, I have two motors that are both NEMA 34 and one is twice the torque of the other due to being twice the stator/frame length, both will mount to a NEMA 34 installation.
Al.

[ger21]

They have two models that are advertised as being on a Nema 34 frame. Are you saying that those would not be drop in replacements for a Nema 34?

No, they are completely different from steppers. Different shaft sizes, different mounting holes, and different body at the shaft end.
The Clearpath Nema 34's are the same as steppers, and can easily be swapped out.