Re: [CAD_CAM_EDM_DRO] Re: mill servo conversion ?'s

carlos_is_dead wrote:

>Something doesn't make sense to me. How can you run 36V motors though
>"fine up to 50V" at 84V?
>

You are not connecting the motor directly to the 84 (or whatever) V
power supply.
It is getting controlled current from the driver, to keep the loop
constraints
satisfied. The 36 V rating is a DC line rating, and means if you
connect the
motor to a stiff 36 V bus, no damage will result. If you connect the
motor to a
stiff 84 V bus, the magnets would likely be damaged. But, the servo amp's
job is to limit the current to the safe value. This requires you to
turn a pot
on the Gecko 320/340 drive, as the pot is shipped at 20. Idiots have to
be protected from not knowing they have to adjust something to prevent
damage.
So, if you use a 36 V power supply, you can't harm the motor, even if
you don't
set the current limit, but you sure won't get full performance.

> Also, how do you know these are 4ohm motors?
>Just from specs that you have, or from some derivation?
>

I examined and measured one.

> When Dave says
>the motors are 12amp max does he mean that is the stall current, or
>does he mean more amperage will damage the motor?
>

More current will damage the magnets. The continuous stall is much lower.
At 12 A, the power dissipation is 576 W, which will fry the motor very
quickly.

> Does this mean, that
>if you buy Dave's setup, it is impossible to get 600in-oz of torque
>even though the motors are rated that high?
>
>

Yes, absolutely. You can only get 9 A, and that is only at stall! So,
that is about
3/4 of rated torque. At 450 RPM, the back EMF is 18 V, and you only
have 18 V
margin available for the I*R drop, you can only get 4.5 A, for 37.5% of
rated torque.

>Honestly, we're not terribly interested in extemely speedy machining
>as we're not a business or anything. We spend a lot more time
>designing than manufacture, its just our size manufacturing
>capabilities are the limiting factor. We are interested in getting the
>most bang for our buck though.
>
>I don't understand how you know the motors I asked about on ebay are
>the exact same motors camtronics sells. What exactly should I be
>looking for in or asking about a motor?
>
>

Ametek does make about 5 different winding configurations on these
motors, but
EVERYBODY selling these motors seems to have that same configuration.

Well, the armature resistance is one parameter, and the Kt and Kv are the
other two parameters you can obtain or measure. The last biggies are the
continuous and peak torque (or current). These are hard to measure without
destroying at least one motor, so you really need to get them from the
manufacturer.
Kt is the torque/amp, Kv is the volts/1000 RPM, and are very important
in figuring
out the belt reduction ratio, if any, that will be needed. To calculate
the electrical
requirements at some specified speed and torque, the current will be
torque/Kt,
and the voltage will be Kv * KRPM + current*resistance. If the power supply
can't deliver this current, then the available torque will drop at this
speed.

>These motors you linked to on ebay are the exact same ones that
>started this thread, that you already said were worse than the ameteks
> camtronics sells. So that is kind of confusing.
>

Yes, sorry. You are right, I'm going in circles here. For Bridgeport
Series I
machines, both of these motors are quite marginal, although either would
work
within its limits. The E660 motor would be much better with Rutex drives at
120-150 V. Since it can take more voltage, it will spin a lot faster.
You then
solve the torque problem with belt reduction. The Gecko's voltage limits
cut into the E660 motor's performance by cutting the speed.

>
>Really, what I'd like to do is buy motors from surplus center,
>encoders from us digital, and find or build a power supply to match. I
>intend to use gecko's so that constrains me with 80V and 20A.
>Comparing to camtronics, they sell 3 motors and the power supply for
>$635. So I'd like to be able to do the same thing for less, or get
>more power for the same amount of money. Also, torque is much more
>important than speed to me. A maximum speed of 50ipm would be plenty.
>We have 2.5:1 gearing on a 4tpi ball screw, so I don't want to run
>over 500rpm at the motor anyway. Our computer can only handle 18khz
>pulsing, so with 320's anything over a 500cpr encoder makes the
>computer the limiting factor anyway.
>
>

Well, given all these constraints, I think the Ametek motors would probably
be a better choice. But, I will tell you that with a 36 V power supply, you
will have some real limitations.

>So, my specifications are coming into focus, and really, I'm just
>writing them out for my own benefit as much as to seek advice on them.
> I'd like to get dc servo motors with as much stall torque, and as
>much torque as possible up to 500rpm. I'd like to get 500cpr encoders
>to match the motors. And I'd like to find or build a power supply to
>work that will work with the geckos.
>
>So, knowing that, how do I choose a motor? What makes a good motor for
>this application? A higher or lower max rpm? Higher or lower voltage?
>Does a higher RPM motor mean there will be higher torque at lower
>RPM's? How do I determine internal resistance? How do I find out stall
>torque (surplus center doesn't usually provide that information)?
>
>

The speed at x voltage is what determines the reduction ratio. The equation
I put together above determines what you can expect given a voltage limit.

>Then, how do I choose a power supply based on those specs? I've heard
>I need about 2/3 the amps of the stall amps of all the motors
>combined. Is that peak amps or sustained? And how do I judge a power
>supplies peak amperage? Can some sustain the peak amperage for longer
>(it seems so) and if so, how long?
>

There are no hard and fast rules, it depends on the safety margins built
into
the supply. Most bulk (unregulated) supplies can withstand a pulse of 200%
rated load for a second or two every minute. But, you can't really
depend on
that.

> What about voltage? How do I choose
>voltage with respect to the motor's voltage and maximizing torque? Is
>higher voltage always better?
>

Up to a point, yes. If it is too high, the triangle-wave current
induced in the
motor by the PWM technique will cause motor heating. You need to run
the equation I did above to figure the worst case voltage demand and then
add a little factor for losses in the PWM amp.

> Also, I've heard that these power
>supplies are rather simple (practically just a transformer, rectifier,
>and capacitor) so is there some way I can build my own?
>
>

Of course, and most people do! There has been a LOT of discussion of this
over the past couple of years.

Jon