[MechanoMan]

Wow, I did not expect this to happen. I believe that's a CNCRouterParts 380oz-in on a CNCRouterParts timing belt drive, 48v G540, across the gantry axis.

After running the router for 10 hrs or so we noticed some non-round holes. I found that gantry axis could move a bit when you push on it, the belt wasn't tight enough. Saw the larger timing pulley shift while the motor pinion didn't. I tightened it a bit but hardly gorilla'ed that tight though.

Was running fine then not-running. Went to check and found this:



Whole output shaft split right at the bearing! Ever seen anything like that??

[gfacer]

I had that happen on a 1kw servo less than 2 years old, on a machine that probably could use 750w or less as the axis I think had the least amount of stress on it.... I was sent a replacement, chalked it up to either manufacturing defect or it was dropped before installed.

[victorofga]

some nema 23 has 3/8 shaft

the picture you posted, shows it broke at a ""shoulder"" possible it was too ""sharp"" in the corner.. might be that was issue, mightbe it was too tight..

yes over tightening can snap this shaft..

thinking on how many step going per minute.. each is a very small knock.. and the millions of knock will snap it..

check also dspi website for a ""softer"" belt, or less width..


check this article..

Failure Analysis: Metal Fatigue of Automatic Machinery - Wafer Test Station


shows same broken surface, and this explain what happened your motor..
of course theres also the chance some factory defect.. but the surface lets guessing it was fatique..

[ger21]

Is that a 1/4" shaft or 3/8" shaft?

[victorofga]

one more tought, can you make the pulley closer to the motor body?

that also would help a lot.. it is visible on the end of the shaft..

[MechanoMan]

That's a 1/4" shaft.

For the CNC Router Parts design, that's where the pinion MUST seat. The motor is on a pretty thick aluminum plate.

[ger21]

That's why they sell those motors with 3/8" shafts now.

[victorofga]

you can fix that on two way..
first put together anyway your router, and make another plate, and make a sink for the stepper, leaving about 5 mm wall
then motor get closer to the other side of the plate..

or cut trough the plate as the motor body, and slip from other side the motor..

================================================== =============

also nema 34 has 1/2 in shaft, and same ounces 34 motor already will work.. wont break..

Dual Shaft Nema 34 CNC Stepper Motor 2.2Nm(312oz.in) 2A 34HS26-2006D|34HS26-2006D|Nema 34 Stepper Motors

[MechanoMan]

Yikes. I have 2 of those 1/4 ones on my X-axis too. I can't afford to replace 'em all.

The CNC RouterParts design has been troubling for me. The rack's pinion is located too far off from the base, IMHO. The offset distance between tension spring pulls it down and where the pinion pushes back does make significant torque on the plate and mount. Seeing how the plate's thickness requiring the pinion be placed too far out on the shaft and capable of breaking the shaft outright is icing on the cake.

Looks like the only place I can get a NEMA23 in 3/8" is CNCRP, and only in 320 oz-in. I'm using the 380 oz-in. I'm well aware a torque number doesn't tell the story as far as performance, but I'm not comfortable with the motor being chosen because it's the only option possible for a 3/8" shaft. KelingInc (automation technology) doesn't have ANY 3/8" options except for the 570oz/in 5A and that's not a good option for a 48v G540.

[victorofga]

well..

thinking on simplicity, just use nema 34..

bad news of this, any modification you need lathe and mill....

or replace the thick aluplate with a steel? what can be definetly thinner?

without tools, you can make only thing split half the belt along, then it more flexible and require less force to tensioning....

with larger motorshaft the pulley also have to have a larger mounting hole..

about design...
you cant expect something like a ford designed.. because development can not exceed the expectable profit..
and in this case looking for their price, they working on very close to the actual material price..

so.. I wouldn't blame on them.. they made possible many folks can afford router, otherwise they cant..

so just be creative and youll find a good solution..

================================================== =====
theres exist the nema 24 standard, and those all have larger shaft..
nema 24 has same mounting pattern than 23

more torque like originally you have

570 Oz-in NEMA 24 Stepper Motor (Dual Shaft) [SM-08] - US$60.00 : Homann Designs!, The preferred CNC Component Supplier

[MechanoMan]

NEMA 34 gets pricey. Not so much that the motors themselves cost more, but a GeckoDrive G540 can't drive the specified currents. You'd need 4x G213v and that's over $500. Plus a much larger power supply with a higher voltage and current.

Steel would likely be thinner to remain practical, and thus less stiff than aluminum.

It'd be great to put the belt pulley and stepper on one side of the plate and the rack-pinion on the other. The stepper would have to be offset on a standoff mount so its pinion isn't on a different side of the main plate, but that does mean the pinion isn't on the far end of the shaft. Then the rack-pinion teeth aren't offset so far out because the 60T timing pulley isn't between the rack-pinion and the plate.

[victorofga]

what I posted 34 that can be driven by the gecko..
motor require 24-60V
and 2.0 A

gecko is rated to 3.5 A and 50V

but again.. I have ideas, but don't know what actually match your imagination..
I just trie dto pointing out, that many option you have..

one of them will work..sure

================================================== ==

placing on opposite side the drive, and other side the pinion.. that would be the best..
but that solution pull up the demand for lathe and mill...

[MechanoMan]

Your link is a NEMA24, not 34. 24 is a real size but

Driving a motor rated for 570 oz-in at 5A with 3.5A is 399 oz-in of torque.

Driving at a lower voltage reduces the high-speed performance. Well more specifically IF your voltage-to-inductance ratio is too low, you can't meet the target current at higher speeds. But adding excess voltage alone doesn't increase speed in the absence of a problem with high inductance, in fact it can make its current rise times excessively fast and make the drive unstable.


What confuses me here- stepper torque/current ratio and inductance go pretty much hand-in-hand. A "high torque" stepper isn't necessarily better because it usually has proportionately higher inductance and can run into problems at high speeds. And that's true across motor frame types.

But here, the 570 oz-in NEMA24 you specified... that specifies only 2.1 mH/phase, a freakishly low inductance. In fact that doesn't seem possible. I thought the current might be from a bipolar-series arrangement while the inductance is from bipolar-parallel. But the info says it's only 4-wire so there's only one option. Unless they screwed up the description somehow.

Hey did you edit your post? I was looking at 570 Oz-in NEMA 24 Stepper Motor (Dual Shaft) [SM-08] - US$60.00 : Homann Designs!, The preferred CNC Component Supplier

As far as the one on your post now: Dual Shaft Nema 34 CNC Stepper Motor 2.2Nm(312oz.in) 2A 34HS26-2006D|34HS26-2006D|Nema 34 Stepper Motors that's a 7mH motor.

According to G540 manual (as well as the higher voltage G213V) you need to use voltage= 32*sqrt (inductance in mh) = 85v which would obviously be a problem as the G540 can only work on 50V max. Actually the motor's spec is only for 60v, it wasn't intended to run all that fast.

[victorofga]

first I posted a nema 34

Dual Shaft Nema 34 CNC Stepper Motor 2.2Nm(312oz.in) 2A 34HS26-2006D|34HS26-2006D|Nema 34 Stepper Motors

regardless of nema size, the amper and volt values are important..
the difference is the really larger shaft

the other I posted a nema 24, and that has the limit over what g540 can perform..

==================================================

im not the electronic guy.. but I was looking for my machine motors, that's why I know a lot stuff around
and the point originally was getting a stronger shaft, avoiding to redesigning the whole router :-)

im not sure how many actual manufacturer exist for steppers.. but I don't think a lot.. so what I can buy at one seller, probably 20 another seller sells same parameter motors..

[MechanoMan]

Yeah but none of the NEMA34 motors are appropriate for a G540. The one you posted would be a poor performer because its inductance is too high for the G540's voltage range, and it doesn't use the G540's current capacity.

[victorofga]

I have xylotex drive and its a small compact system..
it working very well, that's why I purchased a 4 axis to replace the old..
however the new one has a bad resonance...

when I emailed xylotex, the answer was it I s"""because it has Chinese steppers.."""
I told him but the other system don't have resonance..
and since years it is sitting in the box..

so that time I started to looking for, and find beside xylotex a couple of compact stuff exist example the tb6560 on ebay you can see mostly..

I just don't like them..
after that I find the leadshine drivers, and their description just says pretty good things..

1st
since all axis independent, virtually the nema 42 and nema 11 can be in one system

2nd
I don't need to play around with a voltmeter to adjusting mysterious potmeters

3rd
the small sized 4 driver plus 4 motor with powersupply cost same than the xylotex..

4th
the midband resonance filtered out in the leadshine drives..

so from here, makes more sense independent drivers.. and in case, like yours might worth to shift.. a little larger motor don't hurt.. at least you don't loose steps..

I hope you find eventually a sure solution..

[victorofga]

and one more thing, the leadshine drivers reducing current to 60 percent , when motor stops, so motor don't heat up

http://www.leadshine.com/series.aspx...rives&series=M

[MechanoMan]

Well I talked to CNC Router Parts and they offered to swap the remaining 2 intact motors and pinions for the rare NEMA 23 they have with a 3/8" shaft, which seems to be the only one of this type in existence.

There are 2 things which concern me- one, it's only a 320 oz-in motor, and that's the only option. It was cited as a "low inductance motor" but it's actually rather high for its output. With 48v on the G540, the specified inductance would be 2.25 mH. This one is 2.8mH. See that doesn't compare well to KelingInc which has a 2.8mH motor with 381 oz-in rating (but a 1/4" shaft).

At the time I did this, I bought CNCRP's 380 oz-in which I thought were the same thing as Keling, just more expensive. I appear to be mistaken, CNCRP's got a super-high 4.1mH inductance. So it's a better match for inductance to use the 320, but considerably less performance than Keling's with the same inductance.

But the other thing is, the 320's gotta use a 22T pinion instead of a 20T pinion. So it's got 16% less torque (but maybe more speed) and 9% less reduction ratio, which compromises the torque at the rack pinion.

This sounds like a less powerful setup here.

[ger21]

According to G540 manual (as well as the higher voltage G213V) you need to use voltage= 32*sqrt (inductance in mh)

That's actually a maximum recommended voltage.
It also says you can use between 4-20 times the rated voltage.
Using a lower voltage may or may not result in slower speeds.


You really can't compare all these different motors based on specs alone.
Because torque drops off with RPM, you really don't know how much torque each motor will have at any given rpm.

[MechanoMan]

Yeah it's a max- but it's good to be near the max. The "4-20 times the rated voltage" is a guideline for stepper voltage ratings. The stepper voltage is the I*R voltage at the specified current, and not very important in itself. You don't normally size that way, it's a very inaccurate and inconsistent principle since inductance doesn't follow I*R. Well it does say "4x to 20x" so they're honest that it doesn't really give a good suggestion. That could be "9.6v to 48v", which is basically the entire operating range of the G540.

It's not THAT big of a mystery. Higher torque motors have greater inductance, which- IF you have a voltage problem- limits its torque at higher speeds. Otherwise the rolloff isn't all that variable.

When comparing motors of different inductances, it's hard to say. But if they're both 2.8mH, the 380oz-in should outperform the 320oz-in at every speed.