CNC Software CNC Software

Subscribe

 

Categories

Thanks for the Survey Responses!

Feb 23, 2011   //   by   //   Blog  //  6 Comments

I’ve now gotten about 30 survey responses, which is enough to start looking for trends and doing research on some new articles of interest to the readership.

First, some observations about the overall tone and content of the responses:

I really appreciate the constructive view most have taken. With the Internet, and particularly when there is the opportunity to respond anonymously, sometimes you get a lot of hostile or very unconstructive feedback. Not here!

There is one major challenge, but it’s one I’ve known about for a long time–the readership is decidedly bi-modal. What I mean by that is that there is clearly a large group that wants articles that are more “high-end”, “professional”, or “advanced”. There is also a group that wants more “beginner”, “introduction”, “shopmade/homemade”, or “getting started.” I didn’t notice too much in between, and both groups are almost the same size in terms of feedback responses. That’s cool. As I said, I was already aware of it and think it’s very healthy that I don’t go off too far in one direction or the other. Accordingly, I will try hard to publish articles for both audiences.

One other observation–a few bemoan the articles and talk about G-Wizard. Sorry guys, but it pays the bills. You’ll have to learn to cope. Besides which, I do try to write about the products in a way that you could take the lessons discussed and apply them with or without the software.

Second, I wanted to publish some of the specific ideas so you could get a sense for what people are proposing. One thing to be aware of is I try not to write about a subject unless I have direct experience with it and have carefully researched it. So, for example, I had a request to delve into the hand scraping process to produce precision surfaces. This is an area I will get into when I learn how to do it, but I wouldn’t just go read a bunch of articles and then write about it like I had a clue, LOL.

Okay, here are some of the ideas that I think will make good articles, in no particular order:

- “Gcode interpreter for Mariss’s forth coming new drives.” This is a fascinating topic and a broader article about drives, their various formats, motion control boards, and the potential for the two to come together would make a good post.

- “Using a CNC mill as a lathe with gang tooling. How to set that up, tool it, and program it.”

- “DIY cool projects and ideas.” Amen. Always looking for fun quick projects. I was recently casting about for gift-worthy projects that could be CNC’d. I will try to do some of those on my own and write them up in the blog.

- “Scratch built spindles.” It’s coming. When I have finished my mill enclosure, I will probably CNC my lathe next, and then will look into belt drives. Meanwhile, I have been collecting research data like crazy on it.

- “Setting up home and limit switches.” Yep, I’ll be doing this as part of my CNC lathe conversion once I get going. I’ll be sure to detail that all out.

- “Fixturing” Oh yeah, this is one of my favorite topics. If you come across cool fixturing, or tooling that you particularly like for unique workholding, drop me a note!

- “Gear cutting.” Definitely something I will do at some point, if only because I want to add gear calculations to G-Wizard. Gotta build my 4th axis first though!

- “Custom CNC Builds.” You’ll see my lathe conversion before too long (I hope!). In addition, I’ll be on the lookout for cool builds on other sites and try to throw in some pointers here to them. I agree, it’s been a little while for that stuff.

- “Help and Restoration for Users of Old CNC Machines.” There’s some potential for this in conjunction with G-Wizard Editor. One of its missions is to translate g-code as a sort of post-postprocessor. You won’t see something in-depth until I actually get into restoring an old CNC. That’s on my list, but not very high up.

- “High speed machining” You bet, it’s one of my favorite topics. I follow this stuff closely and build it into G-Wizard. When I get my belt drive up and going on the mill, I’ll be showing some more HSM stuff in action too. I think you’ll be surprised at what a small mill can do with a suitably high speed spindle. I also have a series on chatter tee’d up that just awaits final research and software support. I’ve made some posts on Practical Machinist that allude to some of the chatter-related calculations you’ll be able to do. I also got a request to discuss Trochoidal Tool Paths. I’m going to write a little code generator “Wizard” that will do conversational trochoidal slots, just for the heck of it. That will let me play with the parameters. I want to compare Trochoidal with Constant Cutter Engagement paths, for example.

- “Highlight small shops” As many of you know, I do have the Home Shop Hall of Fame. I’d love to add to that. If you have a great small shop, either at hope or a pro shop, drop me a note with some pix. Happy to highlight your shop if it has something unique or interesting about it. The same goes for projects you’re working on, especially if you used G-Wizard to help (natch!).

- “Which cutters?” This is a good topic. The respondent wrote a detailed note about wanting to understand which cutters shops keep in their toolchangers, and how to select the best cutter for a job.

- “Blue Swarf” LOL, love the blue screamin’ meanies! This could also refer to the kinds of HSM things the folks at Blue Swarf are doing. Both are very cool.

- “Mill versus Gantry Router” I need to get more experience with a gantry first. But, the good news is my brother wants one bad, so I’m sure we’ll end up building one before too long. Probably a plasma table too!

- “Ramp Angles” It turns out, there are some upcoming “Mini-Calculators” being added to G-Wizard. We already have one to figure surface finish for ballnoses and turning. I’ll be adding one for figuring feeds and speeds ramping and one for interplated holes.

Phew! That’s a lot of wonderful feedback and great ideas!

Please don’t be offended if I don’t get them all written right away. As I mention, I like to do my homework until I’m very comfortable with a topic before I write it up. Some of these ideas are going to take some more schoolin’ on my part, but that’s always a good thing. In addition to articles I directly research and originate, I love to call attention to other great sites and articles. If you see one you like, drop me a line and I will spread the link love to it. Everyone benefits as my readers get to see a cool article and said cool article gets more traffic.

One more thought: if you think you’d like to try a blog post, why not a Guest Post on CNCCookbook. Send me your idea or article and I’d be happy to take a look. No advertising please, I reserve that for G-Wizard, LOL!

Okay, let’s keep those cards and letters coming. The surveys will be a permanent fixture here, although I may change what they ask from time to time.

 

 

Like what you read on CNCCookbook?

Join 30,000+ Machinists, Designers, Engineers, and Hobbyists!  Get our latest blog posts delivered straight to your email inbox once a week for free. Just enter your email address below:

 

100% Privacy: We will never Spam you!

6 Comments

  • HOME Lever Trick

    Having home built a 3-axis router several years ago, I had used common micro-switches with levers, except on the Z-axis where I directly depressed the nib to actuate.

    When HOMING, the Z-axis was usually within 0.0010″ or less and the worst axis would sometimes be off by 0.0030″ from a previous Home.

    I also tried a test using opto interrupt switching with worse tolerances on repeated Home location.

    I noticed, I was pushing the outer end of the micro-switch lever.

    What I needed was a way to multiply the motion of the axis to actuate the switch faster.

    I moved the contact point where the bed pushes the lever, near to the lever’s pivot, resulting in a great improvement in accuracy.

    My machine uses halfstep driven stepper motors driving a 1/4-20 threaded rod; so, 200 steps x 2 = 400 steps/rpm and 1/20 = 0.05″/rpm; thus each halfstep is 0.000125″ of linear motion.

    The trick is to push on a lever near its pivot and the far end of the lever travels farther than where you push.

    At 18:1 distance ratio, the HOME tolerance will occasionally show 0.0001″, but usually 0, indicating one halfstep of repeat postioning error.

    It is not to difficult to draw in CAD a vertex (pivot center), a baseline to represent the lever, a tiny perpendicular line (halfstep distance), a second line from vertex out 20 times the distance of the contact push point, and measure the displacement (motion multiplied) at the far end of the lever.

    This motion multiplier actuates the switch rapidly, signalling the controller of the Limit condition to Halt. You have reduced the timing signal’s lag and stopped step pulses to the motor sooner; on my machine within a halfstep.

    Several things for me to try are applying this to an opto interrupt and trying this with micro stepping.

    I also stack two switches, the second switch is the Axis overrun switch: halts all motion on all axii to prevent the machine from damaging itself.

    The overrun switch can be actuated by the same lever by removing several thousandth’s of material from the lever end where it contacts the nib; thus, if the home
    switch fails, the axis will only travel a small distance to overrun stop.

  • HOME Lever Trick

    Having home built a 3-axis router several years ago, I had used common micro-switches with levers, except on the Z-axis where I directly depressed the nib to actuate.

    When HOMING, the Z-axis was usually within 0.0010″ or less and the worst axis would sometimes be off by 0.0030″ from a previous Home.

    I also tried a test using opto interrupt switching with worse tolerances on repeated Home location.

    I noticed, I was pushing the outer end of the micro-switch lever.

    What I needed was a way to multiply the motion of the axis to actuate the switch faster.

    I moved the contact point where the bed pushes the lever, near to the lever’s pivot, resulting in a great improvement in accuracy.

    My machine uses halfstep driven stepper motors driving a 1/4-20 threaded rod; so, 200 steps x 2 = 400 steps/rpm and 1/20 = 0.05″/rpm; thus each halfstep is 0.000125″ of linear motion.

    The trick is to push on a lever near its pivot and the far end of the lever travels farther than where you push.

    At 18:1 distance ratio, the HOME tolerance will occasionally show 0.0001″, but usually 0, indicating one halfstep of repeat postioning error.

    It is not to difficult to draw in CAD a vertex (pivot center), a baseline to represent the lever, a tiny perpendicular line (halfstep distance), a second line from vertex out 20 times the distance of the contact push point, and measure the displacement (motion multiplied) at the far end of the lever.

    This motion multiplier actuates the switch rapidly, signalling the controller of the Limit condition to Halt. You have reduced the timing signal’s lag and stopped step pulses to the motor sooner; on my machine within a halfstep.

    Several things for me to try are applying this to an opto interrupt and trying this with micro stepping.

    I also stack two switches, the second switch is the Axis overrun switch: halts all motion on all axii to prevent the machine from damaging itself.

    The overrun switch can be actuated by the same lever by removing several thousandth’s of material from the lever end where it contacts the nib; thus, if the home
    switch fails, the axis will only travel a small distance to overrun stop.

  • HOME Lever Trick Part 2

    In my previous comment, I mentioned the
    original condition of pushing on the micro
    switch lever end. This same condition is
    displayed in the Hossmachine Automatic Tool Changer page, Home Switches: X Axis Limit Switch and Y Axis Limit Switch photos.

    On my machine, the machine would seek toward zero on the axis until it tripped the switch; then it would try to retract a set amount to the home position.

    However, at the reversal of motion after tripping the switch, the retract would occasionally rebreak the contact and cause an error message that the Home position was out of tolerance. Then I jog away and
    restart HOME again for all three axis. Frustrating and with various small position errors.

    I searched for a solution this year resulting in moving the machine’s point of contact to trip the switch near the lever’s pivot.

    An analogy of a windshield wiper will
    illustrate the change I made. The top of the wiper blade sweeps a long arc while the
    bottom end, nearest the pivot, sweeps a
    short arc across a windshield.

    The original condition was contact at the
    top of the wiperblade (Lever) pushing a long distance while the nib on the switch is at the bottom end of the wiperblade nearer the pivot point. In fact the exact opposite of what is desired.

    By moving the point of contact near the lever pivot, the lever pushes the nib at the top end, sweeping a larger arc, and moving faster and farther than the axis motion. This actuates the switch in an abrupt change from closed contacts to open, a clean break in the TTL signal resulting in improved repeated location for the machine.

    If you made the point of contact on the
    lever midway between the pivot and the nib,
    the ratio would be 1:2. By moving closer to
    the pivot, the ratio increases. At 1/4 the
    way from the pivot the ratio is 1:4. Moving
    even closer to the pivot, the ratio increases to 1:16 so the axis moves one unit of distance at contact and the lever end pushing on the nib moves faster and farther at the same time.

    You can also use switches without a built-in lever and make a separate lever to actuate the switch in the same way with machine contact near the pivot and nib depressed at the lever’s far end.

    Makes me wonder how many machines are actually setup like the original condition.

  • HOME Lever Trick Part 2

    In my previous comment, I mentioned the
    original condition of pushing on the micro
    switch lever end. This same condition is
    displayed in the Hossmachine Automatic Tool Changer page, Home Switches: X Axis Limit Switch and Y Axis Limit Switch photos.

    On my machine, the machine would seek toward zero on the axis until it tripped the switch; then it would try to retract a set amount to the home position.

    However, at the reversal of motion after tripping the switch, the retract would occasionally rebreak the contact and cause an error message that the Home position was out of tolerance. Then I jog away and
    restart HOME again for all three axis. Frustrating and with various small position errors.

    I searched for a solution this year resulting in moving the machine’s point of contact to trip the switch near the lever’s pivot.

    An analogy of a windshield wiper will
    illustrate the change I made. The top of the wiper blade sweeps a long arc while the
    bottom end, nearest the pivot, sweeps a
    short arc across a windshield.

    The original condition was contact at the
    top of the wiperblade (Lever) pushing a long distance while the nib on the switch is at the bottom end of the wiperblade nearer the pivot point. In fact the exact opposite of what is desired.

    By moving the point of contact near the lever pivot, the lever pushes the nib at the top end, sweeping a larger arc, and moving faster and farther than the axis motion. This actuates the switch in an abrupt change from closed contacts to open, a clean break in the TTL signal resulting in improved repeated location for the machine.

    If you made the point of contact on the
    lever midway between the pivot and the nib,
    the ratio would be 1:2. By moving closer to
    the pivot, the ratio increases. At 1/4 the
    way from the pivot the ratio is 1:4. Moving
    even closer to the pivot, the ratio increases to 1:16 so the axis moves one unit of distance at contact and the lever end pushing on the nib moves faster and farther at the same time.

    You can also use switches without a built-in lever and make a separate lever to actuate the switch in the same way with machine contact near the pivot and nib depressed at the lever’s far end.

    Makes me wonder how many machines are actually setup like the original condition.

  • HOME Lever Trick Part 3

    My 3-axis router seeks toward axis-LIMIT at 16 inches/minute until it trips the switch. The Signal Generator’s software routine waits for LIMIT “bit” to flip then decelerates the step pulse stream to stop, flips the direction “bit”, accelerates up to speed to retract adding the correct number of pulses to offset the soft HOME position, and decelerates to a stop: axis-HOME.

    Quite a few pulses to add and subtract, so where does the distance error occur?

    If the motor is not missing steps, the deceleration and acceleration have equal steps; then, it might be during the LIMIT switch signal and backlash.

    At 16 inches/minute: 400 halfsteps x 20 revolutions/inch x 16 inches/minute = 8000 x 16 = 128000/(60 seconds/minute) = 2133 halfsteps/second.

    The micro-switch or opto-interrupt construction introduces physical movements to be considered.

    The micro-switch internally has a contact on the end of a sprung lever applying pressure to the other contact surface. To break the connection, the sprung force must be overcome and the inertia of the mechanism changes. Spring force varies from one switch to another. These three factors take time, during which STEP pulses are streamming to the motor. If you push slowly on the switch mechanism, the contact pressure is slowly relieved with lessening electrical flow until no contact and the sprung force is overcome. The transition from high to low voltage setting the LIMIT “bit” is too slow; so varying amounts of STEP pulses pass each attempt to locate the LIMIT. On a dual trace oscillascope, the voltage falloff would be gradual and you could count the streaming pulses passing during this falloff time.

    Similarly, the opto-interrupt detector’s resistance changes the voltage as it is slowly blocked during a constant feed motion of the axis involved.

    Moving the point of contact near the lever pivot, the lever pushes the nib at the top end, sweeping a larger arc, and moving faster and farther than the axis motion.

    At 1:20 (5% of the distance from pivot to nib), if the halfstep movement is 0.000125″, by similar triangles the nib is depressed 0.0025″ at the same time. The velocity at the nib is 20 times greater than the axis velocity.

    On the switches I use, the nib must move 0.030″ = 12 halfsteps to de-contact at 1:20 ratio, leaving 28 more halfsteps (0.070″) before the lever would touch the switch case. 12/2133 = 0.0056 of a second. If you depressed the nib directly (1:1) at 0.000125″/halfstep for 0.0300″ it would take 240 halfsteps to de-contact. If the lever extends beyond the nib a goodly distance, the ratio might be 3:2 or 360 halfsteps to move 0.030″.

    If off by ten halfsteps, a 0.00125″ repeated position error between two HOME attempts. Sometimes the difference is two halfsteps or seven or one or many more depending on the setup. Vibration, dust, contact resistance, backlash compensation, thermal expansion, and wire connections contribute to the soup of uncertainty in an electro-mechacanical interface.

    Actuating the switch in an abrupt change from closed contacts to open in a minimum of time, a sharper voltage dropoff of the TTL signal improves repeated HOME location for the machine.

    Yesterday, I tried a compound arrangement with a second lever at 1:10 ratio pushing the switch lever at 1:10 ratio for a combined total 1:100 ratio. Reduced halfstep count pushed switch’s spring loading before triggering contact separation, similar results of a few halfsteps either side of zero on repeated HOME seeks were produced. Adjusting backlash slightly improved results.

    Keep a list of repeated HOME results with a date, time, and corrective action taken. Plot the results on an X,Y grid or X,y,Z CAD to see where the changes over time are drifting. Plot different date/times with unique symbols to aid in recognizing change.

    You should be able to HOME within 0.0005″ of zero with the same micro switches using machine contact point leverage enhanced precision.

  • HOME Lever Trick Part 3

    My 3-axis router seeks toward axis-LIMIT at 16 inches/minute until it trips the switch. The Signal Generator’s software routine waits for LIMIT “bit” to flip then decelerates the step pulse stream to stop, flips the direction “bit”, accelerates up to speed to retract adding the correct number of pulses to offset the soft HOME position, and decelerates to a stop: axis-HOME.

    Quite a few pulses to add and subtract, so where does the distance error occur?

    If the motor is not missing steps, the deceleration and acceleration have equal steps; then, it might be during the LIMIT switch signal and backlash.

    At 16 inches/minute: 400 halfsteps x 20 revolutions/inch x 16 inches/minute = 8000 x 16 = 128000/(60 seconds/minute) = 2133 halfsteps/second.

    The micro-switch or opto-interrupt construction introduces physical movements to be considered.

    The micro-switch internally has a contact on the end of a sprung lever applying pressure to the other contact surface. To break the connection, the sprung force must be overcome and the inertia of the mechanism changes. Spring force varies from one switch to another. These three factors take time, during which STEP pulses are streamming to the motor. If you push slowly on the switch mechanism, the contact pressure is slowly relieved with lessening electrical flow until no contact and the sprung force is overcome. The transition from high to low voltage setting the LIMIT “bit” is too slow; so varying amounts of STEP pulses pass each attempt to locate the LIMIT. On a dual trace oscillascope, the voltage falloff would be gradual and you could count the streaming pulses passing during this falloff time.

    Similarly, the opto-interrupt detector’s resistance changes the voltage as it is slowly blocked during a constant feed motion of the axis involved.

    Moving the point of contact near the lever pivot, the lever pushes the nib at the top end, sweeping a larger arc, and moving faster and farther than the axis motion.

    At 1:20 (5% of the distance from pivot to nib), if the halfstep movement is 0.000125″, by similar triangles the nib is depressed 0.0025″ at the same time. The velocity at the nib is 20 times greater than the axis velocity.

    On the switches I use, the nib must move 0.030″ = 12 halfsteps to de-contact at 1:20 ratio, leaving 28 more halfsteps (0.070″) before the lever would touch the switch case. 12/2133 = 0.0056 of a second. If you depressed the nib directly (1:1) at 0.000125″/halfstep for 0.0300″ it would take 240 halfsteps to de-contact. If the lever extends beyond the nib a goodly distance, the ratio might be 3:2 or 360 halfsteps to move 0.030″.

    If off by ten halfsteps, a 0.00125″ repeated position error between two HOME attempts. Sometimes the difference is two halfsteps or seven or one or many more depending on the setup. Vibration, dust, contact resistance, backlash compensation, thermal expansion, and wire connections contribute to the soup of uncertainty in an electro-mechacanical interface.

    Actuating the switch in an abrupt change from closed contacts to open in a minimum of time, a sharper voltage dropoff of the TTL signal improves repeated HOME location for the machine.

    Yesterday, I tried a compound arrangement with a second lever at 1:10 ratio pushing the switch lever at 1:10 ratio for a combined total 1:100 ratio. Reduced halfstep count pushed switch’s spring loading before triggering contact separation, similar results of a few halfsteps either side of zero on repeated HOME seeks were produced. Adjusting backlash slightly improved results.

    Keep a list of repeated HOME results with a date, time, and corrective action taken. Plot the results on an X,Y grid or X,y,Z CAD to see where the changes over time are drifting. Plot different date/times with unique symbols to aid in recognizing change.

    You should be able to HOME within 0.0005″ of zero with the same micro switches using machine contact point leverage enhanced precision.

Leave a comment

Home

Software

  GW Calculator

  GW Editor

  Gearotic

  Conversational

  Deals and Steals

CNC Blog

  Software

  Techniques

  Beginner

  Cool

  Projects

 

Cookbooks

     Feeds and Speeds

     G-Code Tutorial

     CNC Machining & Manufacturing

     DIY CNC Cookbook

     CNC Dictionary

CNC Projects

Resources

     Machinist's Search

     Videos

     Online Groups

     Individuals

     Reference Data

     Books

     CNC Dictionary

     Suppliers

     Tool Brands

Workshop

     Hall of Fame

     Organization: Soon!

 

About

     Customers

     Partners

     Our History

     Privacy Policy

 
All material © 2010-2014, CNCCookbook, Inc.