servo motor

[mace]

I found a connection for these servo motors
Would they be any good for cnc router with 3 1/2 porter cable engine I have started to purchasing components for my routers

I need 12 servos for to 5 axis routers and 2 rotary indexers I am building and I think it is important to get all the same servos it
says there 5/8 of a HP

x y and z axis in my design are thk rail and .2 lead ball screws
2 to 1 ratio
4-6th axis will be appox 10 to 1 timing belt set up custom machined components
It all sounds nice to me but this is a big chunk of money for me

I still do not understand what I am looking at in these specs

and i have been reading all the posts on servo motors
And I want Everyones opinon Please


They are $150 with encoders

Reliance Electronics
part # 721-015-0774
model#E712
Torque Constant Kt 37.51 +/- 10% Oz-in/A
Back EMF Constant, Ke 25.0-30.5 V/KRpm within 400mv CW to CCW
Motor Terminal Resistance .75-1.02 Ohms @ 4A.
Nominal Voltage Rating 60V
Nominal Speed Rating 2100 RPM
Maximum Peak Torque 1500 Oz-In.
Maximum Peak Current 44.5A
Maximum Continuous Stall Torque At Max Motor Temp 300 Oz-In.
Maximum Continuous Current 8.9A
Motor Inertia .15 Oz-in-sec
No Load Current at 50v .90A Max
Maximum Winding Temp 155 deg C
Shaft End Play .001" Max with 20lb load
Four replaceable brushes.
Motor Body 5.675" Long, 4.0" Diameter
Shaft 1.3" Long, 0.5" Diameter.
1.5" Diameter Positioning Boss, 4 10-32nd Mounting Holes on 2.5" Diameter.



Thanks to all With knowledge that Help me Decide if these are a worthy investment
Mace

[balsaman]

look like lots of power to me...you could email the specs to geckodrive to see what he thinks. Will you be using Gecko320's? What is the encoder CPR?

Eric

[mace]

He gives me a choice of these So I think the 500cpr is the ones that i should buy But i dont Know!

I am thinking rutex r90h due to higher 40 amp rating.

U.S. Digital encoders in 250cpr (1,000ppr with quadrature) or 500cpr (2,000ppr with quadrature) for an additional $45 per motor.


Mace

[balsaman]

You could go rutex but I don't think you will need the 40 amps. My motors are 1.1 amps cont. running on 75 volts and I have enough power....but that is just me.

I would get the 250 cpr encoders.

Eric

[mace]

Is There a advantage of getting the 250 cpr encoder ?
So far the ball screws that I bought are .0001 accuracy

I think the servos get more accurate with a greater number of ppr Is this a true statement?

I am just curious why you would pick the lesser of the two if there is no extra charge!(is there something i am missing ?)

Thanks Mace

[ballendo]

Garrett,

The reason many folks go with the lower resolution encoders is to keep from having "empty" resolution, which inhibits top-end speed.

Empty rez simply means you can't REALLY get as small a unit move as you "think" you can. And confusion of resolution and accuracy is epidemic, IMO, among new cnc'ers...

Aaron of industrialhobbies.com recently posted a bit to his website called balancing the servo and computer, which explains in a bit more detail.

Now you have ground ballscrews, so the screw CAN "get there"

Can your drive get there too? And if your drive CAN; will your computer controller be able to keep up? Or will the system bog down speed-wise, because the control loop cannot be executed/updated often enough at higher speeds?

Hope this helps,

Ballendo

P.S. With screws of .0001 accuracy, to get their full value; you might want to use an encoder which gives about 40K steps per inch. Using Mach2 as an example, you can see that having only 45K steps/secon available means that your top speed will be limited to 67-1/2 IPM... (Even at 20K SPI, which is the minimum to use to achieve the accuracy your screws are capable of, you're only gonna traverse at 135IPM. BUT...

This assumes that EVERYTHING ELSE is similarly accurate! Structure, temperature control, both environmental AND machine component--means the screws are gonna heat up due to friction--, screw end mounts, etc.

I've seen VERY few machines home built that REALLY can get plus or minus .001! (Think about it. even in jobber machine shops, "typical" part tolerance is plus/minus .005!!!)

It CAN be done, but it's not gonna happen with aluminum structure(at all), or welded tubing(of the size commonly seen in DIY).

FWIW, with your screws, and since I know you have some background in machines, i'd use a 500 or 1000 encoder. times 4 gives .0005 or .00025 ,which is still probably much better than the machine structure will achieve.

Originally posted by mace Is There a advantage of getting the 250 cpr encoder ? So far the ball screws that I bought are .0001 accuracy I think the servos get more accurate with a greater number of ppr Is this a true statement? I am just curious why you would pick the lesser of the two if there is no extra charge!(is there something i am missing ?) Thanks Mace

[mace]

Thank you very much ballendo!!!
That is what i was missing

Mace

[HuFlungDung]

Wish I worked in a typical job shop. The only thing I am happy with producing at +/- .005" is the chip thickness

[sdwilson]

i ran across these on ebay the guys says he has hundreds of these, whats your thought on these for motors.

there like 459.00 buy it now. are these motors worth the money

my machine im trying to build is more for metal cutting


Here are the details:

Reliance Electric DC Servo motor

Reliance Servo Products

NEMA 34 Frame

Model 0660

Back EMF Constant, ke = 65 V/kRPM

Torque Constant, kt = 5.5 lb-In/Amp WOW!

That's PER Amp of motor current.

At just 10 amps, that's 55 Lb-In = 880 Oz-In.

Nominal Speed Rating: 1750 RPM

Nominal Voltage Rating: 120 Volts

De-mag Current: 40 Amps

Motor Body Diameter: 3.38"

Motor Body Length: 6.75"

Shaft Diameter: 0.375" x 1.13" Long

Mounting per NEMA 34

(4) 0.225" Holes on 2.75" Square Centers

Integral Encoder: Agilent / Hewlett Packard

500 Line yielding 2000 Counts/Rev. WOW!

Encoder is mounted under protective steel cover on tail end of motor.

[ballendo]

Originally posted by HuFlungDung Wish I worked in a typical job shop. The only thing I am happy with producing at +/- .005" is the chip thickness

Hu,

So what do you work to, on a feature to feature basis, "typically"?

(For those who may not know, feature to feature means from holes to hole, or wall to wall, or... Yes, the hole itself WILL typically be held to a tighter tolerance by a jobber shop.)

We were talking about the inherent accuracy of a given machine, remember?

So on that note, let's insert a bit of REALITY amongst all the "claims" for plus/minus .001...

The thermal coefficient of expansion for 6061-t6 aluminum from 20 to 100 degrees celsius (approx 40 degrees to 212 degrees F, or inclusive of our normal shop environment) is...

13.1x10-6, or .0000131 PER INCH, PER DEGREE Fahrenheit.

So an aluminum structural rail of a homemade CNC machine which is only 24 inches long will grow and shrink .0003144 in length FOR EVERY DEGREE OF TEMPERATURE CHANGE.

At 10 degrees, that's .003! (And I'm willing to bet that most shops and garages where these things are operating undergo a MUCH greater temperature variation )

Now let's add another bit of "truth" to this discussion. I see many machines with steppers mounted directly to the framework... Steppers get hot. REAL hot. Aluminum is a great conductor of heat. Do I really need to say more?!?!?

Now I'm not saying this to discourage anyone from making a machine out of aluminum. But let's be mindful that we live in the physical reality of the world, no matter that we might like to believe otherwise...

Hope this helps,

Ballendo

P.S. On a 4x8 foot machine, which often has 10 foot rails, the CTE is .016 for each 10 degrees F of change. That's the machine structure itself!

P.P.S. But I'm using steel, you say? The CTE of 1018 steel is 6.39x10-6, or about 48% of aluminum. So it will move less, but IT STILL MOVES SIGNIFICANTLY. In the specific case of this thread (pun intended), If Garrett's ground ballscrews are 24 inches long, they will move .0001536 PER DEGREE of temp change.

So in our 10 degree shop temperature swing example, they move about 1 and 1/2 thousandths...

Taking it to a bridgeport mill with more like 48 inches of screw length on x, and you're talking about .003 inaccuracy of the screw itself for only 10 degrees of change. And let's not forget that the screw is heating up due to friction. (So that 10 degrees does NOT even need to be "felt" in the shop air to make this example valid.)

Another factor that works IN our favor is that most of our feature to feature distances are NOT 24 inches. (I want to keep this REAL.) But most of the temp changes involved are MUCH greater than 10 degrees!

Now if you're cutting steel, and the machine has steel screws, then some of this cancels out. But think about what happens when you place a block of aluminum on a bridgeport mill...

And once you've done that, I'm pretty sure you'll agree that most of the machines here just DON'T achieve the ACCURACIES claimed for them... (At least not without the actions of a dedicated AND INFORMED machinist(which you may be, Hu Flung.), willing to adjust for the physics involved!

But I never see any posts on THAT topic...

[HomeCNC]

SDWilson,

I have also seen thoes. They look to be very nice servo motors for a full size mill.

[turmite]

Originally posted by ballendo [FOR EVERY DEGREE OF TEMPERATURE CHANGE. At 10 degrees, that's .003! (And I'm willing to bet that most shops and garages where these things are operating undergo a MUCH greater temperature variation )

ballendo, I was thinking more along the lines of 60 deg. change from summer to winter. But then again I am cutting wood and I have no idea what the change for wood wood be. I just couldn't resist that one.

I make rifle stocks and pistol grips and fought the accuracy bug the first few yrs of business. I found out if I can maintain +/- .015
it's good enough. I bet you guys would like those kind of tolerences huh??

Mike