Counter balancing Z axis..

[adam_m]

OK, at least I'm not totally nuts.

I was also looking at a quill return spring... same question there, how to ID the correct application..

[louieatienza]

g540.... sorry missed that.

[adam_m]

I'm getting myself off track as this round type of spring although might work has some design and machining time involved... I was hoping for quicker/easier solution.

The quickest I think is either the air strut or gas strut... I'm leaning towards a gas strut.

Adam,

[louieatienza]

I'm getting myself off track as this round type of spring although might work has some design and machining time involved... I was hoping for quicker/easier solution.

The quickest I think is either the air strut or gas strut... I'm leaning towards a gas strut.

Adam,

The quickest thing would be a counterweight and pulley. You can get an LM rail and block to restrict the counterweight 's movement...

[Need TECH Help!]

There are some reducible gas struts at McMaster.

[adam_m]

Any of those spiral springs will also be limited by cycles at some point... that could be a problem.

Jim Dawson, what Air spring are you currently using in your setup. I'm curios to see how that setup works..

[Jim Dawson]

My air spring is on my mill for backlash control. I used a 1.562 dia x 6 inch stroke air cylinder and have it attached to the quill with a cable. On a router you could attach the air cylinder directly to the Z axis slide. The system consists of an air cylinder, a regulator, an air tank, and a few fittings. I supply the regulator with shop air, and it is regulated down to about 30 PSI to run the air spring. The system uses almost no air because it is a closed system. The only air loss, which is minimal is just from the normal very minor leakage.

Take a look here and scroll down to the last two pictures

http://www.cnczone.com/forums/knee-v...-software.html

[NIC 77]

On a router you could attach the air cylinder directly to the Z axis slide.

That is exactly what I have been thinking of over the past several hours. Thanks to 109jb's comments.

P1V1 = P2V2. Thanks for the reminder on the equations 109jb!

It appears to me that a 5 gallon tank is not required to use a gas cylinder! The higher the pressure you use, the less volume is required.

The only air loss, which is minimal is just from the normal very minor leakage.

That's something we need to pin down. How much leakage? For a gas strut, zero leakage. For an air cylinder, how much? Could I run a machine for a week on a closed system without topping it up from a compressor?

So for adam_m's example. 8 inches of movement. A gas cylinder with an id of 16mm (available on EBay), pie*r^2 = 0.312 in^2. A typical home air compressor can do around 120 lbs per in^2, you could charge that up to 96 PSI and get 30 lbs of force at the top end.

So with a tank of 5 inches in diameter and 6 inches in length, that's 117.8 in^3. So total volume is (0.312 x 8) + 117.8 = 120.3 in^3.

Pressure1 x Volume1 = P2V2, 96 x 120.3 = ?? x 117.8

?? = 98 PSI, which gives 98 x 0.312 = 30.6 lbs of force at the bottom end of the stroke. So yes, you could do this without the 5 gallon air tank if you change the operating pressure. A bottle from a fire extinguisher or a small propane tank (would have to check the operating pressures of those items?) directly attached to the Z should do the trick if you want to use a gas cylinder. Or am I missing something?

Why did you choose the value of 30 to 40 PSI? Very important to know.

Of course, a gas spring without dampening could do this too, but not be adjustable. The less expense gas struts don't even mention dampening, but I assume that they are all dampened to a large degree.

So feel free to point out errors in my reasoning, I won't be offended.

Take a look here and scroll down to the last two pictures

http://www.cnczone.com/forums/knee-v...-software.html[/QUOTE]

Nice work Sir! I've read that build thread but didn't equate your name to it. I was considering a mill conversion, but it just didn't have the amount of movement I need. So I ended up on another CNC build.

adam_m,

I've had a look at your motor specs. I can't find a torque vs. speed graph. But, for a 570 oz-in stepper, assuming the 0.5 was the lead and not the diameter, I can't see you having less than 100 lbs of thrust available at 100 IPM, ignoring inertial effects (I still need to do those for ballscrews in my calculation sheet). And your acceleration is at a measly 0.038 G's. I'm assuming that your Z axis (the part that moves) is around 30 lbs from your comments?

So here is what I would do in your shoes:

Try increasing the acceleration to 0.1G and see what happens. Then try a few more acceleration settings above or below what you had set before and see what happens.

Then try a new motor and drive. Change the drive and motor you use between z and x for example. Change the settings in mach 3 to match between the axis and whatever different voltages if you have different motor sizes on z and x. oooh, nice if they are the same size, so you can do this. Yep, guess you can't do this if they are not the same size as it won't fit the mount. So forget what I said about voltage.

30 lbs is not heavy for a z axis, so I seriously doubt that this problem will go away with the addition of a counterbalance, especially if you have run this machine before and did not experience the same problem.

It was working before? How long has it run in the past without problems? We can continue to discuss counterbalancing, but in the mean time, let's assume that this is not the root cause of your problems.

[louieatienza]

adam_m,

I've had a look at your motor specs. I can't find a torque vs. speed graph. But, for a 570 oz-in stepper, assuming the 0.5 was the lead and not the diameter, I can't see you having less than 100 lbs of thrust available at 100 IPM, ignoring inertial effects (I still need to do those for ballscrews in my calculation sheet). And your acceleration is at a measly 0.038 G's. I'm assuming that your Z axis (the part that moves) is around 30 lbs from your comments?

So here is what I would do in your shoes:

Try increasing the acceleration to 0.1G and see what happens. Then try a few more acceleration settings above or below what you had set before and see what happens.

Then try a new motor and drive. Change the drive and motor you use between z and x for example. Change the settings in mach 3 to match between the axis and whatever different voltages if you have different motor sizes on z and x. oooh, nice if they are the same size, so you can do this. Yep, guess you can't do this if they are not the same size as it won't fit the mount. So forget what I said about voltage.

30 lbs is not heavy for a z axis, so I seriously doubt that this problem will go away with the addition of a counterbalance, especially if you have run this machine before and did not experience the same problem.

It was working before? How long has it run in the past without problems? We can continue to discuss counterbalancing, but in the mean time, let's assume that this is not the root cause of your problems.

He has a Gecko G540 and his steps/in is 10160, so at 2000 steps/rev that's 5.08rev/in, he has a 5mm lead ballscrew. Which with the 570in-oz stepper should move 30lbs easily. I moved a 20lb gantry with a 1" lead ACME screw for years with a 425in-oz stepper.

[ger21]

I've had a look at your motor specs. I can't find a torque vs. speed graph. But, for a 570 oz-in stepper,

You're looking at the wrong motor. The 570oz motor has an "M" on the end of the number. He has the 381oz motor.


http://www.automationtechnologiesinc...orquecurve.pdf

[louieatienza]

That motor is rated at 5A... since you can only run it max 3.5A with the G540 you have 3.5/5*570 or 399in-oz stall torque, or your running at 70% the stepper's potential torque. Though at 350rpm, according to Keling, the motor is still putting out 2.75N-m (389in-oz) at about 5A... at 3.5A that's about 272in-oz, still plenty enough I think to get that Z going.

"EDIT" - posted this while Gerry responded...

[louieatienza]

You're looking at the wrong motor. The 570oz motor has an "M" on the end of the number. He has the 381oz motor.


http://www.automationtechnologiesinc...orquecurve.pdf

Well, there's the problem... was actually looking at NIC77's last post... though even a 380in-oz motor should move that Z....

[109jb]

If you have an air cylinder with a good piston, the leakage will be essentially zero. It should work for an indefinite period without leakage. If it leaks, replace the piston seal, usually just an o-ring.

A gas strut has an orfice/hole in the piston and the hole size will determine the amount of damping the strut has. Damping is one of the traits of a strut that is considered beneficial in its typical application. You want the hatck on your car to open slowly and close slowly instead of slamming up and down. Just because of the nature of things this dampening is most noticeable at the extreme ends of travel. On a typical CNC this damping won't be too detrimental, but I could see it being detrimental on a really high speed one. An Air cylinder will have essentially no dampening.

There was some mention of using a smaller reservoir than 5 gallons, but I personally would use larger and about 60-80 psi.

I looked up your build, and if I am right, this is a picture of your Z-axis. If I am wrong, please correct me.:


From other information in this thread, here is what I see:

Linear guides, 5mm ballscrew, 381 oz-in 3.5A stepper, gecko 540. I didn't see anywhere that you mentioned the voltage you are running to the steppers but it should be about 50VDC.

From this, I would say you have something else going on. I can run the Z-axis of my G0704 that runs on dovetail slides (more friction), with cheap KL6050 drivers and 570 oz-in motors at 200 IPM all day long with no counterbalance. I can push it up to 250 IPM but that is too close to the lost steps area for my taste. You setup should run at least 100 IPM with no problem as i'm sure your router is lighter than my cast iron head with heavy DC motor on top, and has much less resistance with the linear rails.. Are you sure you have the G540 set up to provide maximum 3.5A current, and are you running close to 50V?

[NIC 77]

You're looking at the wrong motor. The 570oz motor has an "M" on the end of the number. He has the 381oz motor.

http://www.automationtechnologiesinc...orquecurve.pdf

good catch!

He has a Gecko G540 and his steps/in is 10160, so at 2000 steps/rev that's 5.08rev/in, he has a 5mm lead ballscrew.

another good catch!

http://www.automationtechnologiesinc...lat-381-oz-in/

so it looks like this requires 3.5Amps. G540 max 3.5 Amps, so that's good. Power supply is 48 VDC.

http://www.nookindustries.com/Linear...e_Calculations

So looking at the basic nook equation Td = (F x Ph x 10^-3)/(2 x pie x n)

We get:



with a 50% efficiency and not including inertia, as I have already said, I have yet to to those for ballscrews.

The 5mm lead makes a big difference. There's no way in my opinion that he should have less than 100 lbs thrust at 100 IPM with that motor, all things considered.

[louieatienza]

If you have an air cylinder with a good piston, the leakage will be essentially zero. It should work for an indefinite period without leakage. If it leaks, replace the piston seal, usually just an o-ring.

A gas strut has an orfice/hole in the piston and the hole size will determine the amount of damping the strut has. Damping is one of the traits of a strut that is considered beneficial in its typical application. You want the hatck on your car to open slowly and close slowly instead of slamming up and down. Just because of the nature of things this dampening is most noticeable at the extreme ends of travel. On a typical CNC this damping won't be too detrimental, but I could see it being detrimental on a really high speed one. An Air cylinder will have essentially no dampening.

There was some mention of using a smaller reservoir than 5 gallons, but I personally would use larger and about 60-80 psi.

I looked up your build, and if I am right, this is a picture of your Z-axis. If I am wrong, please correct me.:


From other information in this thread, here is what I see:

Linear guides, 5mm ballscrew, 381 oz-in 3.5A stepper, gecko 540. I didn't see anywhere that you mentioned the voltage you are running to the steppers but it should be about 50VDC.

From this, I would say you have something else going on. I can run the Z-axis of my G0704 that runs on dovetail slides (more friction), with cheap KL6050 drivers and 570 oz-in motors at 200 IPM all day long with no counterbalance. I can push it up to 250 IPM but that is too close to the lost steps area for my taste. You setup should run at least 100 IPM with no problem as i'm sure your router is lighter than my cast iron head with heavy DC motor on top, and has much less resistance with the linear rails.. Are you sure you have the G540 set up to provide maximum 3.5A current, and are you running close to 50V?

Is that a 2-start or 5-start ACME screw? If so, then the steps/in are wrong. Second, you're running that screw directly to the motor via a coupler? Is it fully locked in at the non-driven end? If not, steppers can have some axial play, and that may show up as lost steps if you're plowing into material and that foce exceeds that which is holding the rotor in place. Plus that coupling acts as a big spring - it can be prone to wind-up.

[NIC 77]

There was some mention of using a smaller reservoir than 5 gallons, but I personally would use larger and about 60-80 psi.

Are you saying you would use larger than 5 gallons or that you would drop the pressure to 60-80 PSI? Why may I ask? Even at 70 PSI, a small air tank attached to the Z is totally doable, even for my build, where I don't care if I am off by 10 or 20 lbs.

Please explain your reasoning, I very much would like to hear what you have to say on the subject.

[louieatienza]

I suspect now that's a 2-start ACME, which should be exactly 10000 steps/inch, and you've compensated because there's some compliance somewhere, which made you come up with 10166? Or a 5-start which should give you exactly 4000 steps/in (as in your original build thread? Changed screw?

[ger21]

Are you still using this nut?

http://www.cnczone.com/forums/attach...2&d=1383179219

And are you lubricating the screw? If you don't have perfect alignment with that nut, you can get a lot of binding. If you pull the motor off, is the screw easy to turn with your fingers? It should be.

[NIC 77]

Oh hey adam_m,

You fly RC? me too.

[louieatienza]

Are you still using this nut?

http://www.cnczone.com/forums/attach...2&d=1383179219

And are you lubricating the screw? If you don't have perfect alignment with that nut, you can get a lot of binding. If you pull the motor off, is the screw easy to turn with your fingers? It should be.

Looks like the tensioner is missing too...

Adam did mention spindle drop when the system is off, possibly this is why there are lost steps, because the missing tensioner is causing lash? I know even though I had an 8-start screw (1tpi) there was still enough tension to hold about 20lb Z carriage assembly.