[IQChallenged]

I have been using my CNC router for a couple of years with great satisfaction despite some blatant violations of sensible design. The most obvious short cut is the absence of any kind of gear reduction. In all three axes I have a small pinion on the motor shaft meshed directly with a rack. My motors are probably bigger than they need to be at 640oz. Everything works well despite the aforementioned configuration that certainly isn't a good recipe for resolution and quiet operation. So I want to take off enough time from project building to add a gear reduction. I'm especially eager to get rid of the buzzing that goes along with the motors having to run so slowly. I usually run them at about 60ipm. The buzzing disappears for the most part at 420ipm, a factor of 7x. Before I start ordering pulleys and belts I wanted to see if anyone knew of any adjustments that might help so that I didn't have to make such a drastic change. I know there are some screws on the Geckos that can be turned but I don't dare do so without knowing what they do. I was really looking forward to the Mach seminar but unfortunately it got canceled. Any help will be appreciated.

www.ProjectCNC.wordpress.com

[Charles Lessig]

IQC,

If you are using stepper motors, too little inertia may cause
vibration at certain speeds. Some motors wanted an external
inertia equal to that of the motor's rotor. A mechanical damper
is another solution to the problem. It is a small flywheel coupled
to the motor shaft by friction. It limits acceleration a little but
it is very simple.

Here is a drawing on page 5,

http://www.zilog.com/docs/z8encore/appnotes/an0128.pdf

Best regards, Charlie

[harryn]

Hello IQ - I have read your build info several times and actually am planning to build my Z axis with a direct drive pinion on rack setup. I could not tell from your web site the specs for the rack and pinion, but I assume pretty small.

Some versions of the motor have fairly high inductance, which makes switching at high frequency tricky. I wonder if you turned down the microstepping to more like 10 or 100X if it reduces the buzz ?

The damper is a really good thing to try. There are some examples of just how well it works in the stepper motor section. (videos). I am guessing they are youtube videos.

OK, here is one thread on it, but there are several if you search the stepper section on dampers. http://cnczone.com/forums/showthread...ghlight=damper

[IQChallenged]

Hello IQ - I have read your build info several times and actually am planning to build my Z axis with a direct drive pinion on rack setup. I could not tell from your web site the specs for the rack and pinion, but I assume pretty small.

Some versions of the motor have fairly high inductance, which makes switching at high frequency tricky. I wonder if you turned down the microstepping to more like 10 or 100X if it reduces the buzz ?

The damper is a really good thing to try. There are some examples of just how well it works in the stepper motor section. (videos). I am guessing they are youtube videos.

OK, here is one thread on it, but there are several if you search the stepper section on dampers. http://cnczone.com/forums/showthread...ghlight=damper

Thanks Charlie and Harryn. I'll get some detail on my Z and post it.

[IQChallenged]

Both rack and pinion (6325K13 GEAR, 6295K16 RACK) came from

http://www.mcmaster.com/#

[Mike Everman]

Nice work, IQ. Do you not have a counterbalance cylinder or spring? If not, that addition would improve your situation immensely.

[IQChallenged]

Nice work, IQ. Do you not have a counterbalance cylinder or spring? If not, that addition would improve your situation immensely.

No, I don't. And that reminds me. When I put up those pictures last nite I should've included a warning. If you use R&P for your zed-- you should know that when you turn off the power the stepper will immediately go into neutral and the sled will come crashing down to the table. One must jog down close to the table and then shut it off. As far as springs or cylinders I'm assuming you're talking about making it more zero gravity to reduce work on the motors. Yes? No?

From an operational point of view this thing works great as is. The only 2 things I don't like is the resonance from lugging them so slowly (about 60ipm), and knowing that I could do more precision work with more resolution. It's a big Harley with nothing in the gear box but a fifth gear. Well- second gear.

When I drew the thing up I had never seen a CNC with R&P and didn't know if anyone had ever done it. Since then I've learned that they're commonly used on routers. I'm guessing that they use powerful enough spindles that low speed is not needed and big enough projects so that resolution is not an issue. And of course- much, much cheaper.

[Crevice Reamer]

what Geckos are you running?

CR.

[IQChallenged]

what Geckos are you running?

CR.

201s.

[Crevice Reamer]

The 201 HAS mid band resonance compensation--unless it's been turned off.

ADJUST: This trimpot adjusts the motor for the smoothest possible low-speed operation. Set the motor speed to about 1/4 revolution per second, (15 RPM) and then turn the trimpot until a distinct null is noted in the motor’s vibration. This will result in the most
even microstep placement for a given motor and power supply voltage.

http://geckodrive.com/upload/G201-REV-16-MANUAL.pdf

If this doesn't help, you should call Gecko support:

Contact Technical Support

If you have a technical question and would like to contact technical support,
you can do so by the following means:

Telephone:

(714) 832 - 8874

Fax:

(714) 832 - 8082

Email:

[email protected]

CR.

[IQChallenged]

The 201 HAS mid band resonance compensation--unless it's been turned off.

Thanks reamer. How would it get turned off? I'd be tickled ----less to find that I could make a lot of it go away. But I'll have to find out exactly how to do it. I know there's a little screw on the drivers. I guess that's what you turn. I'll also have to find out what a null is :-). I was planning on going to the Mach seminar but it got canceled.

[Crevice Reamer]

Null means the motor gets as quiet as it can be. Yes, turn the screw inside that little hole. If you call Gecko, they will walk you through it.

CR.

[IQChallenged]

Null means the motor gets as quiet as it can be. Yes, turn the screw inside that little hole. If you call Gecko, they will walk you through it.

CR.

I tried it. Unfortunately they were already optimal. And I really shouldn't complain. And I'm not. Anyone reading this and wondering whether or not they can get satisfactory performance in making a unit like might should know that they can even without a gear reduction. And for a lot of folks that makes CNC attainable whereas the leadscrew option (because of the higher cost) might not be.

[harryn]

Hi - your rack and pinion selection looks fine - at least to me. (16 tooth pinion, 16 pitch, 14.5 degree PA)

I am considering R/P on my Z axis and there is not enough room to make the pinion side spring loaded - so I am considering to spring load the rack. Maybe that will work for you ? Without the spring load or some other way to force the r/p to mesh all along the way, it gets dicey.

I realize that everyone likes to do the ultimate in micro stepping, esp. when you have not geared down, but let's look at the numbers.

1 revolution of your pinion = Pi x D = 3.14 x 1 inch = 3.14 inches.

With no microstepping, you have 3.14 inch / 200 steps = 0.016 inches for each stepper motor step. (about 1/64 th inch)

In wood, do you really need much better cutting size control than 8 - 10 X better than this ? I am guessing that your microstep setting is like 100 - 250 ? More microsteps risks more buzz, esp. if the motor inductance is large - like many more powerful motors have.

Somebody - please double check my math here.

In other words, try turning the microstepping either off or way down and see if it makes any difference.

The damper is also a very good path.

BTW - very nice and clean build.

[Crevice Reamer]

The G201 is fixed at 10 microsteps, and this cannot be changed.

What size Power supply are you using? What Current set resistors? Are the drives heat sinked?

Here are the internal switch settings:

HEATSINKING: The G201 needs heatsinking for current settings greater than 3 amps. The case temperature (measured on the
bottom plate) should not exceed 70 degrees C, and for best life should be kept to 50 degrees or less. Use heatsink compound
between the G201 and the heatsink.
CAUTION! Current settings above 3 Amps without a heatsink will result in damage to the G201.
The drive must be heatsinked to a piece of aluminum, preferably with fins and a fan to increase heat dissipation and surface area.
Do not screw the drives directly to the door of your control cabinet, as this will typically not provide adequate heatsinking
properties.
OPTION JUMER BLOCK: The G201 has four settable options. These options are enabled/disabled via the jumper block shunt
settings. See the diagram below for how to set the shunts. The four options are described below.
(1) AUTO CURRENT REDUCTION: The G201 reduces motor phase current to 33% of the set value when the motor is stopped.
If enabled, this reduction occurs 1 second after the last step pulse is sent to the drive. Auto current reduction can be enabled or
disabled.
(2) REDUCED CURRENT RANGE: In addition to the normal current range (1A to 7A), the G201 can also operate over a reduced
current range (0.3A to 2A). This range is used for motor phase currents of less than 1A. Auto current reduction is not available for
this reduced current range.
(3) NEMA-42 MOTOR: Use this option if the drive is to be used with NEMA-42 frame size motors. This option adjusts the midband
compensation circuit for optimum performance with this size motor.
(4) MID-BAND COMPENSATION DISABLE: Some step pulse sources generate very “ragged” timing. If the pulse-to-pulse period
varies by more than +/-30%, the drive may have problems counting that step pulse. This option disables the mid-band
compensation circuit, making the drive insensitive to this timing restriction. As a consequence, the motor may exhibit mid-band
instability problems.

There is a diagram of how these should be set and how they can be changed, on page 3:

http://geckodrive.com/upload/G201-REV-16-MANUAL.pdf

CR.

[IQChallenged]

"I am considering to spring load the rack. Maybe that will work for you?"

I kicked it around for a while back when I was making up my mind which way to go. I finally decided it would be to hard to build. But you may figure out a good way to do it and then I'll copy your design.

"I realize that everyone likes to do the ultimate in micro stepping..."

Before reading your post I didn't even know what micro-stepping was (is). Still don't. Apparently you're referring to a way of getting more resolution. As hard as it is for me to believe- the thing cuts very small pieces with remarkable precision as is. I guess the biggest reason for wanting to gear down is to let the motors run faster which makes them much quieter and smoother.

"The damper is also a very good path."

I should probably give that a try. I've never actually seen a clear cut set of plans for one. I'll check.

"BTW - very nice and clean build."

Thank you

[IQChallenged]

The G201 is fixed at 10 microsteps, and this cannot be changed. There is a diagram of how these should be set and how they can be changed, on page 3: http://geckodrive.com/upload/G201-REV-16-MANUAL.pdf CR.

It all seems to check out. I think I'm optimized.

[Torchhead]

It's going to be difficult getting the vibration out of the motion. The direct coupling gives you LOTS of speed (upwards of 2000 IPM of your motor could hack it) but 1/3 (based on 1" dia pinion) the torque and 1/3 the resolution. You could see as poor as .015 resolution and trying to lift that router (I have that exact one myself). You have to rotate the motor slow to hit moderate 100 IPM rates and that causes vibrations.

The Gecko resonance compensation can only do so much. The 203 has an improved compensation and I have a conveyor oven (for SMT soldering) that is run with a stepper than has to spin very slow. Even with the Gecko adjusted it still vibrates the whole machine from the slow motion.

I normally recommend that R & P systems have belt reduction between the Pinon and the motor. Even 3:1 will gain you back the lost torque and resolution and also provide a level of dampening between the motor and the load. That still leaves you with speed you can't use but cancels out the lost torque and poor resolution.

I won't go into it here but microstepping cannot be counted on for increasing resolution or accuracy. It will help at some speeds, and be of little help at others. It's to smooth out the motion and help with resonance.

1000 IPM rapids SOUND good on a spec sheet but running a machine with those speeds can be a scary experience. One wrong move and somebody/something gets hurt! We run a servo machine with major belt reduction to the R & P (10:1) and a heavy gantry (twin gantry motors) and It will easily do 600 to 800 IPM with gobs of torque. I have it detuned to 350 IPM so I can get my old frail butt out of the way if something goes wrong! It's also hard on the motion components to run them around and do jackrabbit stops and starts from high speed.

What you need on a router table: Torque and resolution. It's hard to get either with direct coupled pinons.


The other issue is one you mentioned: The Z drops like a rock when the motors are not locked (turn off) That is because the weight of the Z load has mechanical advantage driving the pinion with the rack. That's okay for horz axis but not where gravity gets involved.

TOM CAUDLE
www.CandCNC.com

[Mike Everman]

Good points all, Tom, and well put!
More and more I see applications for my ServoBelt (which has a pinion pitch diameter of .940") needing gearheads. I need to hurry up and get the big belt version going.
Direct pinion drive is fantastic for pick and place.

[IQChallenged]

Direct pinion drive is fantastic for pick and place.

Can you define what is meant by "pick and place"?