Hello Guest it is August 19, 2019, 06:45:18 AM

Author Topic: CNC'ing both quill and knee  (Read 15982 times)

0 Members and 1 Guest are viewing this topic.

Re: CNC'ing both quill and knee
« Reply #30 on: February 19, 2009, 11:58:54 AM »
I seen your posts and that sounds like great performance.   Unfortunately, you are using servos and mine is a stepper system.    I am not sure how servo vs stepper rated torque would correspond.   Servos like more rpm's so that would explain your 4.8:1 drive ratio.    Right now, I am leaning towards 1810ozin stepper and a 3:1 belt drive.   I'd like 4:1 but using a online calculator the smaller pulley would not have enough teeth engaged at any reasonable center distance.   I could use idler pulleys to increase the belt wrap on the small pulley but I want to keep it simple so 3:1 at about 12-14 inches center distance between the pulleys is about as good as I can get.   Its looking like air springs are going to be necessary to get any reasonable speed.   

As I mentioned, the knee is only going to move during tool offsets at tool change time so speed isn't terribly important.
However.... going from a 1" long endmill to something like a 6" long drill and chuck would mean a 5" move which at 15ipm would take 20 seconds....... pretty slow and that would no doubt get aggrivating fast!   30-40ipm would be quite satisfactory.

You're not going to get high speed from a stepper pushing that kind of load, no matter how you gear it.  The high-RPM torque fall-off will kill you.  So, either live with low top speed, or go to a servo (which is about the same cost as a large stepper anyway).  First step in any case should be to measure how much torque is required to move your knee, pad that a bit, and you can calculate how well any motor will work, and what speed you should be able to get.  Mine was about 40-45 in-lbs.

There's something wrong with your calculations.  I have 4.8:1 reduction (48:1 overall, including the bevel gears and leadscrew) using XL belts - 72 tooth on the screw, 15 on the motor, with only about a 9" center distance.  This gives more than 6 teeth engaged on the motor, which is all that's required.  I also have two 200# gas springs.

Regards,
Ray L.

I have stepper drives for the 4 axis I now have so will stick with steppers for that reason.

I was using a online calculator from Stock Drive Products (SDP) website.   They have a nice calculator for drive ratios, center distances, etc.    When the tooth engagement gets too small the number turn red.   Anything much over 3:1 shows a problem with engagement on the small pulley.   I believe the engagement at 4:1 was indeed 6 teeth at 12 inch center distance but that came up in red which is supposed to mean it is not recomended.   

Oh yea, I know what may be making the difference...... I am using 3/8 pitch L series belts in 3/4 width.   A smaller pitch may be different.   I used 3/8 pitch for my x and y because of the stepper power and the loads driven.
Re: CNC'ing both quill and knee
« Reply #31 on: February 19, 2009, 12:16:59 PM »
Yes, belt pitch makes a big difference.  As I said, I used XL, which is 5-pitch.  Again, with a stepper, you'll be hard-pressed to get better than *maybe* 25 IPM.  A large stepper is NOT going to go fast (maybe 1000RPM at the very best), and torque falls off rapidly with speed, so you'll likely become torque-limited long before you reach that max speed.  With a 3:1 reduction, 750 RPM would give you 25IPM.  If you have to go to 2:1, you'll likely be looking at 15-20 IPM max.

Regards,
Ray L.
Regards,
Ray L.
Re: CNC'ing both quill and knee
« Reply #32 on: February 19, 2009, 09:25:29 PM »
Yes, belt pitch makes a big difference.  As I said, I used XL, which is 5-pitch.  Again, with a stepper, you'll be hard-pressed to get better than *maybe* 25 IPM.  A large stepper is NOT going to go fast (maybe 1000RPM at the very best), and torque falls off rapidly with speed, so you'll likely become torque-limited long before you reach that max speed.  With a 3:1 reduction, 750 RPM would give you 25IPM.  If you have to go to 2:1, you'll likely be looking at 15-20 IPM max.


Regards,
Ray L.

Its not a pretty picture, that I agree.   I see you used a pair of 200lb gas springs which seems to be common.  I was looking at them tonight at McMaster and see they also make them 225lb and 250lb fairly cheap.   

I searched around and found this weight info over on the BP yahoo forum:
> for a standard J-head the weights are:
>
> Head 200#
> Ram Adapter 90#
> Ram 218#
> Turret 214#
> Column 715#
> Knee 257#
> Saddle 142#
> 48" Table 365#
> Total 2101#

Taking a wild guess that a 42 inch table is 200lbs, that makes 599 lbs the motor is moving, lets add another 100 for a kurt vise, the one shot oiler etc and we have 700 lbs jumping up and down!    I think I am going to go with the 250lb gas springs after finding this information.  Any wonder some people find the knee speeds slow!!!

I don't think I want to use a belt as small as a XL for moving such weight.   Granted there is the screw and gears but its still a awful lot of force on such a tiny belt especially considering the accelleration.
« Last Edit: February 19, 2009, 09:42:47 PM by Sparky_NY »
Re: CNC'ing both quill and knee
« Reply #33 on: February 19, 2009, 09:34:10 PM »
Yes, belt pitch makes a big difference.  As I said, I used XL, which is 5-pitch.  Again, with a stepper, you'll be hard-pressed to get better than *maybe* 25 IPM.  A large stepper is NOT going to go fast (maybe 1000RPM at the very best), and torque falls off rapidly with speed, so you'll likely become torque-limited long before you reach that max speed.  With a 3:1 reduction, 750 RPM would give you 25IPM.  If you have to go to 2:1, you'll likely be looking at 15-20 IPM max.


Regards,
Ray L.

Its not a pretty picture, that I agree.   I see you used a pair of 200lb gas springs which seems to be common.  I was looking at them tonight at McMaster and see they also make them 225lb and 250lb fairly cheap.    Do the 200lb ones fully counterbalance the knee?   What do you think of using stronger ones?    Who knows what a 42 inch BP table, saddle and knee weigh other than a guess?

If anything, I have a bit too much.  It now requires less torque to move the knee up than down.

Regards,
Ray L.
Regards,
Ray L.
Re: CNC'ing both quill and knee
« Reply #34 on: February 19, 2009, 10:07:21 PM »
Yes, belt pitch makes a big difference.  As I said, I used XL, which is 5-pitch.  Again, with a stepper, you'll be hard-pressed to get better than *maybe* 25 IPM.  A large stepper is NOT going to go fast (maybe 1000RPM at the very best), and torque falls off rapidly with speed, so you'll likely become torque-limited long before you reach that max speed.  With a 3:1 reduction, 750 RPM would give you 25IPM.  If you have to go to 2:1, you'll likely be looking at 15-20 IPM max.


Regards,
Ray L.

Its not a pretty picture, that I agree.   I see you used a pair of 200lb gas springs which seems to be common.  I was looking at them tonight at McMaster and see they also make them 225lb and 250lb fairly cheap.    Do the 200lb ones fully counterbalance the knee?   What do you think of using stronger ones?    Who knows what a 42 inch BP table, saddle and knee weigh other than a guess?

If anything, I have a bit too much.  It now requires less torque to move the knee up than down.

Regards,
Ray L.

Huh, now that is a bit strange.   Those weights seem to be reasonable, as well as the total machine weight.  In theory if the weight is totally "absorbed" by the gas springs the knee should move with almost no effort except the friction of the ways / screw and gears.   I am thinking of those gas springs on the big hatch back cars, the hatch weighs 200lbs or so and lifts/closes with one finger.

I also in my ventures seen a power drive kit for the knee, it was spec'd at 152 inlbs which comes out to about 2500oz-in, right around the numbers that have been being thrown around considering the belt reductions.

Offline Hood

*
  •  25,849 25,849
  • Carnoustie, Scotland
    • View Profile
Re: CNC'ing both quill and knee
« Reply #35 on: February 20, 2009, 03:05:41 AM »

I don't think I want to use a belt as small as a XL for moving such weight.   Granted there is the screw and gears but its still a awful lot of force on such a tiny belt especially considering the accelleration.

You would be very surprised what these belts can handle, I know I was. Granted I am using a GT3 belt from gates rather than the standard timing belt but I was surprised when Gates told me that for the 17Hp servo I have on my spindle that the 20mm wide 8mm pitch belt would actually be a 60% over engineered solution with the pulleys I was using.This belt has replaced 6 x B section V belts that the lathe used with the induction motor that was originally on it.
  Best thing to do is go to Gates website and register and then download their sizing software.
Hood
« Last Edit: February 20, 2009, 03:07:21 AM by Hood »
Re: CNC'ing both quill and knee
« Reply #36 on: February 20, 2009, 11:06:58 AM »

I don't think I want to use a belt as small as a XL for moving such weight.   Granted there is the screw and gears but its still a awful lot of force on such a tiny belt especially considering the accelleration.

You would be very surprised what these belts can handle, I know I was. Granted I am using a GT3 belt from gates rather than the standard timing belt but I was surprised when Gates told me that for the 17Hp servo I have on my spindle that the 20mm wide 8mm pitch belt would actually be a 60% over engineered solution with the pulleys I was using.This belt has replaced 6 x B section V belts that the lathe used with the induction motor that was originally on it.
  Best thing to do is go to Gates website and register and then download their sizing software.
Hood

And the "weight" being moved is totally irrelevent.  All that matters is the torque at the motor.  If the belt is going to fail, it's going to fail at the motor.  If an XL belt is sufficient to move X & Y, then it's quite capable of moving Z, using the same motor.  Don't design by "feel".  Read the manufacturers recommendations, and the actual ratings for the belts.  It's all in the specs.

Regards,
Ray L.
Regards,
Ray L.
Re: CNC'ing both quill and knee
« Reply #37 on: February 20, 2009, 12:31:22 PM »

I don't think I want to use a belt as small as a XL for moving such weight.   Granted there is the screw and gears but its still a awful lot of force on such a tiny belt especially considering the accelleration.

You would be very surprised what these belts can handle, I know I was. Granted I am using a GT3 belt from gates rather than the standard timing belt but I was surprised when Gates told me that for the 17Hp servo I have on my spindle that the 20mm wide 8mm pitch belt would actually be a 60% over engineered solution with the pulleys I was using.This belt has replaced 6 x B section V belts that the lathe used with the induction motor that was originally on it.
  Best thing to do is go to Gates website and register and then download their sizing software.
Hood

And the "weight" being moved is totally irrelevent.  All that matters is the torque at the motor.  If the belt is going to fail, it's going to fail at the motor.  If an XL belt is sufficient to move X & Y, then it's quite capable of moving Z, using the same motor.  Don't design by "feel".  Read the manufacturers recommendations, and the actual ratings for the belts.  It's all in the specs.

Regards,
Ray L.


When doing my bridgeport conversion last winter I looked at every commercial conversion out there (and homebrew).
All the pics of the commercial ones appeared to use 3/8 pitch belt drives for the x and y.  The knee stepper will be double the torque of the x or y so you are correct the torque transmitted is much higher due to the motor.

I wonder what size drive belts bridgeport used on the boss mills?   I'd be shocked to find they are as small as XL's.

Offline dja

*
  •  18 18
    • View Profile
Re: CNC'ing both quill and knee
« Reply #38 on: January 22, 2010, 06:57:49 PM »
did you need to use a backlash in mach because of the gears on the knee or do you have a digital readout
Re: CNC'ing both quill and knee
« Reply #39 on: January 22, 2010, 08:22:45 PM »
did you need to use a backlash in mach because of the gears on the knee or do you have a digital readout

Now that I have a quill drive, the knee gets used only for applying tool length compensation, so backlash is not an issue.  All knee moves are upward, even if that means first moving below the target position, then coming back up.  This ensures the (considerably) backlash is always taken up, so no Mach3 backlash compensation is required.  Even before I had the quill drive, the knee backlash was not a problem, since I do only 2.5D milling, so moves to cutting depth were always upwards moves.  I got perfectly good accuracy despite the considerable backlash.

Regards,
Ray L.
Regards,
Ray L.