Re:305oz Steppers! More torque, Scotty!!!! (Shes' breaking up, I cannah hold'er)

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "ballendo" <ballendo@...> wrote:
>
> >In CCED, "Dave Rigotti" <drigotti@> wrote:
> >HobbyCNC is pleased to announce that our CNC "Packages" are now
> >available with 305oz steppers! We also have them "ala carte"
>
> Hello,
>
> Another increase in stepper motor torque! Sounds great!
>
> (But there's a fly in the ointment, and nobody seems to be telling
> you about it. IMO it's time somebody did...)
>
> Let's review a couple things about stepper motors, and let's look at
> some published facts about motors likely to be VERY similar to the
> ones just announced, and the earlier/formerly "greatly increased
> performance" motors. (Quotes just above are for MY emphasis, NOT a
> direct quote from someone else.)
>
> I don't know for sure where Dave R gets his motors; but I do know
> that most of us selling motors with these specs are using the same
> motors... Made by the MS motor company in China, and many in the USA
> will source them from John at KelingINC.net
>
> So, let's review:
>
> Most of the time the Stepper torque published and mentioned is
> HOLDING torque. It's the biggest number, so it sounds the best for
> marketing. It's also the way steppers have been identified for
> decades. (Probably due to the previous sentence!<G>) BUT...
>
> It's important to realise... The motor will ONLY have this much
> torque when it is STOPPED. As soon as it starts to turn, the torque
> goes DOWN. FOR ALL stepper MOTORS. Regardless of drive type (bipolar
> vs. unipolar)
>
> How FAR down, and how FAST the torque drops, and along what "shape"
> or "Curve" depends upon some things about the motor, AND some things
> about the drive and power supply.
>
> Every motor is a generator when it turns. This is what causes the
> motor to lose torque as it goes faster. The important criteria for
> this are the motor power supply voltage, and the motor coil
> inductance. An additional factor is the drive type, unipolar vs.
> Bipolar.
>
> So let's look at the numbers for the Motors Keling sells as having
> the specs listed on Daves HCNC website. (Where the new 305 oz.in.
> motors are listed as 6v,2.0A) Here are links to both 305's and the
> 200's as formerly sold by hobbyCNC (and still available from
> http://www.cncresource.com ):
>
> http://www.kelinginc.net/KL23H286-20-08B.pdf (2.15Nm, 305 oz.in.)
>
> http://www.kelinginc.net/KL23H276-30-8B.pdf (1.4 Nm, 200 oz.in.)
>
> Note: A Nm is a Newton-meter, which equates to about 141 oz.in.
>
> (Again I'll mention that I don't know exactly where Dave gets his
> motors; so these may not be the IDENTICAL motors he's selling. But
> I'd bet that if not, they're VERY close!)
>
> The first thing to notice is that the coil inductance for the 305 is
> over THREE TIMES as high (6.8mH) as for the 200 (2.2mH)...
>
> Note: mH is milliHenry's, a measure of inductance. Which is the
> motor coil's resistance to changing current direction.
>
> What this means (that matters to us in the DIY-CNC arena) is that
> the 305's are going to be a better generator of electricity than the
> 200's. Which is NOT good when you're looking for torque at high
> speed. Especially when you're limited in your ability to increase
> power supply voltage to compensate. More on that in a minute.
>
> So the motor turns and as it turns it generates a voltage. This
> voltage is called back EMF (ElectroMotive Force; which is the old
> name for voltage. And it's why Ohm's law is E=IR instead of V=IR).
> Back EMF basically subtracts from the power supply voltage.
> Which means that the motor will behave as if it is being powered by
> a supply voltage equal to the original power supply MINUS the back
> EMF.
>
> Keling doesn't show torque speed curves for all its motors and the
> ones it DOES show have the bipolar torque curve (because bipolar
> drive are more efficient, so that curve makes the motors LOOK
> better). Point is, we can't "see" the torque curve. But we CAN see
> that the back EMF will be higher than the 200's (due to the 6.8mH
> vs. 2.2mH that we CAN see on the spec sheets.)
>
> Translated into CNC-speak, the 305's will lose torque very quickly
> compared to the 200's as speed goes up. In fact, I wouldn't be
> surprised to find that the 200's outperform the 305's at "typical"
> speeds used in most DIY-CNC projects... (just a guess right now;
> I'll know for sure very soon!<G>)
>
> Let's get to the other factors in play here:
>
> Power supply voltage. If we can raise the power supply voltage, we
> can push the 305's to higher performance. And this is true for ANY
> stepper, up to a point of diminishing returns--and motor heating--
> that Mariss of Gecko has empirically determined to be about 20-25x
> motor nameplate voltage.
>
> Side Note: It is STILL BEST to always use the LOWEST power supply
> voltage which will give you the results you NEED. Sizing your PS to
> the max that the drive can handle "just cuz" is bad engineering. And
> that can lead to other unseen and unconsdered problems that you'll
> bat your head against later...
>
> The 305's have a 6 volt nameplate rating. The 200's are rated for
> 2.76V. (using ohms law to solve since this spec isn't on the linked
> sheet. Note that many of the "200's" imported into the USA were
> marked as having 1 ohm coils; in that case the motor volts will be
> 3V.)
>
> So as long as we use drives which have about double the voltage
> capability we may actually see the performance increase with the
> 305's. BUT...
>
> HobbyCNC drives are based on the Sanken SLA7062 (old revs) and
> SLA7068 (pro) driver chips. These chips have a MAX voltage rating of
> 44VDC. (We use these in our CNCResource drives as well; they're good
> chips.) Mariss has pointed out that a unipolar drive semiconductor
> has to be able to withstand DOUBLE the supply voltage. So these
> chips are probably using a 100v architecture, with the de-rating to
> 44V acounting for die variation and reliability improvement.
>
> HobbyCNC also recommends a 24VAC transformer secondary for the power
> supply. This will provide approximately 34VDC when rectified and
> filtered as suggested at the HCNC site. If you're going to buy these
> 305's, it would be a GOOD idea to increase your power supply voltage
> as much as you dare towards that 44VDC limit. If you do that the
> back EMF will be subtracted from this larger "number", and the
> torque will be carried out further along the torque-speed curve.
>
> Wait there's more...
>
> The 305's have a case thats almost 3-1/2 inches (88mm) long. The
> 200's are 3 inches (76mm). That in itself "may" not be too bad; see
> my P.S. following this post!<G> BUT...
>
> The ROTOR INSIDE the motor is ALSO longer. More mass. Put simply,
> slower. (all else being equal. Which I've pointed out IS the likely
> case where the power supply voltage is concerned due to limits of
> the driver chip used.) Another thing which doesn't get much "press"
> her in the build-it-yourself CNC groips is soemthing called
> mechanical impedance matching. Which can be thought of simply as:
> putting a v8 on a bicycle "might" give you problems! You've got to
> match the driven load to the driving motor (AND its defining
> parameters like drive type, power supply voltage and such!) to get
> anywhere near the deisred and expected performance.
>
> And there's another thing called detent torque that we shold
> mention. This is the effect of the reidual magnetic force acting on
> the mechanical shape of the rotor and stator stampings inside the
> motor. It is higher for the longer motors, and its effect comes
> right off the top of your torque expectation. It could be likened to
> a magnetic "friction" that must alwyas be overcome. Shorter motors
> usually have smaller detent torque.
>
> These tow factors are part of why you've read of drives being
> adjusted for larger motor types. Their mass and mechanics are
> different and this directly affect the results.
> You won't see this mentioned when only holding oz-in is being held
> up as THE decision maker!
>
> The point of all this is to say: Don't just look at the biggest
> number oz.in. quoted to decide your purchase!!! There's a LOT more
> to good CNC machine design than simply "bigger is better".
>
> And I'm taking the time to type this because when someone comes out
> and says only, "Bigger motors, get 'em here!" MY job gets harder.
>
> Because then I! have to explain all the things above to OUR
> CNCresource customer who's hearing "bigger, better, bigger, better"
> in his or her mind due to the one-sided push of "Performance" sales
> being done by others in the DIY-CNC arena...
>
> Now let's be fair and explain exactly WHERE the 305's are gonna be a
> good choice. Where you need LOW speed torque. Please keep in mind
> that this is low speed AT THE MOTOR. You can have a very high speed
> machine that has low motor speed. Belt and rack/pinion driven axes
> are good examples. The tradeoff there is step size/ machine
> resolution.
>
> Folks who've seen my writings over the years know that I often talk
> about the "balance" of DIY-CNC design. It's fun for me to see the
> design evolution of many "home builds" over on "the zone"; where
> after MANY iterations they finally empirically arrive at what a
> balanced design approach will give in the first or second
> pass...
>
> Finally, let's mention that the bipolar drives offered by Xylotex
> trade increased performance due to drive type for lower power supply
> capability and lower amps/coil specs, compared to HCNC and
> CNCresource.com driver boxes. FWIW, Xylotex are great drives, and
> Jeff supports them well.
> The net effect is that all three perform similarly with similar
> sized motors. (I've run all three in our lab and the empirical
> results support this statement. There ARE ways to skew the results
> in favor of the unipolar drives; which is why we're currently
> offering the SLA based unipolar drives at CNCresource. We'll be
> adding some bipolar drive options this summer.
>
> Anyways, hope this helps,
>
> Ballendo
>
> P.S. About those increasing motor lengths... thise of us in steppers
> used to expect that the motor body length would fall into roughly
> three lengths. We called these single, double and triple stack
> motors. This due to the fact that inside the rotor WAS divided into
> 1,2,or 3 distinct areas, or "stacks" of laminations. When the new
> range of Chinese motors came out; the old rules were blurred.
>
> Chinese mfrs. were "stretching" the "old" definition of what single
> stack and double stack meant... And pushing the holding torque
> number above all else...
> The first wave was sort of a stack and a half; then double stacks
> like the 200's. Now we're seeing triple stack motors being used in
> what were formerly "single stack" places. Because everybody's going
> after the "torque" by chasing the single HOLDING torque number, as
> explained above.
>
> Anyways, when I designed the Wood Duck CNC routers, we were using a
> 2-1/4" body length motor. Allowing for the "growth" I'd been seeing;
> I designed the machine to be capable of handling 2-3/4" bodied
> motors. Then the 3" motors hit the market. And everybody felt
> they "had to" have those (I've since re-designed to accomodate
> these, but I'm NOT going to keep adding just to "keep up with the
> Jones's"!<G> Especially when the "Jones's" is based on a partially
> explained truth... (We mfrs. are not operating in a vacuum, and we
> have to either educate or explain our competitors moves to our
> potential customers. Sometimes it's easier to just go with the flow.
> But there comes a time when the truth needs to be spoken.
>
> There's more to good CNC machine design than maximising oz-in in the
> motor spec! (FAR Better to maximise oz-in in the axis travels!) The
> two are NOT the same thing.
>