[skibuilder]

I’m betting that this thickness planer mod has been discussed somewhere before on this forum, but I’ve done numerous searches both here and on the web, and haven’t come up with any hits. I have a 20 inch width capacity thickness planer that I would like to convert to CNC. I make snow skis as a hobby and I’d like to use the planer to ‘profile’ the wooden cores for the skis. These cores are about 72 inches long and vary in thickness from about 3/32 at the ends, to 1/2 inch at the center. Up until now, I’ve been using a router bridge set-up, but I’m just about finished with my first CNC router that I’d planned to be taking over the task of profiling. But, I’ve been bitten bad by the CNC bug. I’ve been thinking that a thickness planer would do the job much faster if I could accurately control the variable cutting depth. So, my plan is to remove the current hand wheel depth adjustment on my planer and replace it with a fairly large stepper motor. The G-code should be pretty easy to write because there are only about 6 points along the longitudinal axis of the core where the thickness changes.

Since the planer already has its own in-feed system, the only CNC control I would need would be the depth of cut (I’ll call it the Z axis), so a single axis driver is all that would be necessary… I think. The only part that I’m not too confident about is whether or not the control software (Mach3, most likely) will accept data regarding feed rate (I’ll call this one the X axis) when that data is not input as step and direction (or encoder) data for a stepper (or servo) motor. Basically, the feed rate data would need to be input somewhere into the control software in terms of ‘feet per minute’ or ‘inches per second’. Does anyone have any thoughts, suggestions, or prior experience with this type of software ‘work-around’? Where in the software control panel would I enter this information?

I plan to use an optical sensor (similar to that used to sense when a piece of paper enters a computer printer) to inform the control software when a core blank has been fed into the planer.

Thanks in advance,

Dana

[kenrik]

Since the planer already has its own in-feed system, the only CNC control I would need would be the depth of cut (I’ll call it the Z axis), so a single axis driver is all that would be necessary… I think. The only part that I’m not too confident about is whether or not the control software (Mach3, most likely) will accept data regarding feed rate (I’ll call this one the X axis) when that data is not input as step and direction (or encoder) data for a stepper (or servo) motor. Basically, the feed rate data would need to be input somewhere into the control software in terms of ‘feet per minute’ or ‘inches per second’. Does anyone have any thoughts, suggestions, or prior experience with this type of software ‘work-around’? Where in the software control panel would I enter this information?

I plan to use an optical sensor (similar to that used to sense when a piece of paper enters a computer printer) to inform the control software when a core blank has been fed into the planer.

Thanks in advance,

Dana

Sounds like a very interesting project. Have you figured out how you will control the powered in feed and out feed rollers to follow the variation in the profile Heights as you are feeding the material through the planer ?

[Greolt]

I'm having trouble imagining how you will get the feed rollers and cutter head to move up and down independently.

Greg

[zenbot]

just a thought. you could replace the feed roller motor with a stepper motor. Then you just need to figure out the steps per inch of travel on the surface of the feed rollers. You could completely control the cutting action of the machine this way.

[skibuilder]

Hey guys,

Thanks for the responses. I re-read my original post and can see that I may not have been very clear with my writing. I'll try to elaborate a bit more clearly. The ski cores blanks, before thickness profiling, are a consistent thickness of about 9/16 inch (about 72 inches long and 6 inches wide). After profiling, they are 3/32 thick for about 12 inches at each end and 1/2 inch thick for about 8 inches in the center. There is a consistent taper between the end flats and the center flat.

With my factory-stock thickness planer, I can do a fairly crappy job of profiling the cores just by manually turning the hand crank clockwise and counter-clockwise (raising and lowering the feed rollers and cutter head) at the appropriate points along the length of the core blank (tapers up, then, tapers down). My planer will only allow a 1/8 inch maximum depth of cut, so I have to make about 3 passes for a finished core. If I could control the raising and lowering of the feed rollers and cutter head more accurately, like with a stepper motor driven from some appropriate G-code, I think the planer could do a real fine job of profiling.

Yes, zenbot, I thought about changing the drive motor to a stepper motor, but the present drive motor (a big ol' 5 horse belt driving motor) powers both the (geared together) cutter head and the feed rollers. I'll give it some more thought, though... there may be some way to detach the feed rollers from the gearing system, and then drive the feed rollers with a stepper. Driving the feed with a stepper would just about solve my dilemma... well, maybe not... there's still the task of getting the optical sensor to tell the 'Z-axis' when to initiate the cutting depth values.

I may have to end up building a custom thickness planer that has stepper driven feed rollers on the bottom and floating pressure rollers on top, then a separate stepper positioned overhead cutter-head.

I'd still like to try to get the current planer to work, though. There may be something in the Mach3 software that will allow me to enter the stock machine feed rate speed for the 'x-axis' value.

Thanks again for responding to the thread.

Dana

[Madclicker]

You have 2 axes to control. Why even try to control one without controlling the other?

This is just a simple problem of controlling the feed speed and the height of cut.

Sounds like you just don't want to build the dedicated machine that you know you need to.

[bluejay_ca]

If you can do it by hand then it is at least possible. You are assuming that feed is constant. An input for presence of part entering planer would start program and continue to provide signal for repeated passes.



Personally I'd use a plc to monitor feed rate and a stand alone servo/stepper drive with position inputs. Once the program was finished the planer would open up for the next board.

[Adamj12b]

Could you post some more information about the planer? Maybe some pictures of the drive system and stuff. It certainly sounds possible to accomplish. Also the feed should be the same with a 5Hp motor, That size planer is hard to bog down.

[skibuilder]

Yes, Madclicker, you may be correct regarding the dedicated machine. I spend a lot of time on a hobby skibuilder's forum where a lot of folks are trying to figure out how to get their thickness planers to function reasonably well as profilers. I was just hoping to come up with a planer mod that a number of others, besides myself, could benefit from. I knew that the thickness planer cnc conversion could be a long shot.

Bluejay-ca, could you elaborate a little on what a 'plc' is. Also, what do you mean when you refer to a 'stand alone servo/stepper drive'? Whether I do or don't end up building a dedicated profiling machine, I have a feeling that I could benefit from understanding more about your suggestions. Regarding a modified feed system, it takes an awful lot of force to get a board to feed against the planer's cutting head (even with very sharp blades), so if I did use a servo/stepper to drive the feed system, I'd probably have to incorporate some element of gear reduction. I'm not too concerned about speed at this point. I'd just like to better consistency and accuracy.

Adamj12b, I have a 20 inch Grizzly planer. It's about 20 years old, but still looks just like the model in the current catalog. You are correct in that it does take quite a bit to bog down that big ol' machine. The in-feed rollers have a pretty aggressive bite to them, so the feed rate is pretty reliable. I took the gear housing cover off today, and noted that it is a chain drive. I don't know why I never noticed it before, but the feed speed selector has a "idle' position where the feed rollers don't turn. I don't think that will help me any though, because if the feed rollers aren't turning, I don't think I could feed a piece of wood past them, even if I were to figure out a way to use a stepper or servo to manage the feed aspect.

I'm feeling pretty much split 50/50 right now regarding whether I think the conversion might work vs. needing to build a whole new machine. A new machine would be a pretty hefty project... maybe a bit much for 'hobby' status. So, I'll probably keep dinking with the conversion idea for awhile. Ha, I need another project like I need another hole in my head. I haven't slept much since I discovered cnczone.

Thanks again for the feedback.

Dana

[VWSatOz]

What about a simple copy mill? Place your workpiece on a long base board with a template on one side or both sides set below the cutters lowest point, fit cam follers near to the cutter head and just feed your work through, the cutter head will smoothly feed up & down following the templates, then use twinned-drive down feed screws to adjust the depth of each pass. Sounds such a simple profile why bother with CNC

[skibuilder]

Hi VWSatOZ,

I hadn't heard of a 'copy mill' before reading your post. It perked my interest so I did a google search, but the only thing I could come up with was cnc based copy mills. Do you happen to know of a link that would have a picture of a machine similar to what you describe? From your description, I think that I get the general idea. I'm currently using a router 'bridge' set-up to profile the ski cores. It uses template rails (one on each side of the core) to guides the depth of cut of the router. The router has a 1 1/2 inch 'fly cutter' type bit, so the actual process only takes about 5 minutes, which is fast enough that I can easily live with that part of the process. The main reason for wanting to go to a cnc based machine is that every pair of skis varies a bit in the thickness of one or more areas of the core. Even a 1/32 inch change in thickness in a particular area of the core can significantly change the performance characteristics of the ski, so I might have 30 or 40 basic core profiles that I'd like to be able to draw from. I have tons of templates already and they're kind of a pain to keep track of. It now takes me a good part of a day to make a new set of template rails. If I had a cnc programmable thickness planer, I could draw the profile up in CAD in about 5 minutes, then let the machine cut the profile. I could certainly use my almost finished cnc router to to profile the core blanks, but the cores get so thin (tips and tails) when profiling, that I'd have build a vacuum table to hold them down because the router tends to suck up and destroy them. In my experiments with my planer, it seems to handle the thin sections pretty well (I run them through on a shuttle board).

Thanks for taking the time to share your thoughts.

Dana

[Oldmanandhistoy]

Can you not use an encoder on the feed rollers to monitor position?

John

[VWSatOz]

It sounds like you already have a form of copy mill with your router following a pair of templates rails, but I didn't realise you needed so many profiles to draw from your library. I can believe that the thickness is critical in the core profile as you are trying to create the perfect bending beam to flex in exactly the way you want it to, bend section modulus Z of a bending beam varies by the square of the thickness dimension, so , and this would always vary with the weight of the skier, the length of the ski, maybe even the softness of the snow, and how you envisage what would be the way you want the ski bottom & edges and curve shape to behave/become for different requirements of the skier in a turn, straight downhill, or on a jump etc. So... I would now agree that you continue to follow that plan for your building a cnc mill, but with the ability to fit a HORIZONTAL head atachment to the Z axis of your machine, to able to be quickly be fitted on 2dowels with 4 screws etc... mounting onto a flat vertical face of your Zaxis carriage, in place of the usual vertical router spindle.... 5minute changeover time for the two different cutter heads should be easy to achieve. It would be good if you can keep both horiz. & vert. choices available from your cnc as this gives you unlimited possibilities for any uses & any future designs. ie you MAY find it that your largish fly cutter that you now use is the way to go as you are then not restrained to have just a flat top surface core, and the cutter shallow scolluped surface shape result is no disadvantage. I have just been trying an old Makita 6inch wide hand held electric planer I bought from a garage sale, quite a monster to hold, but somthing like this might have a wide enough cut for your skis. This must be the biggest model they ever made, but could easily be set up at an angle with the front toe plate removed, and mounted on your gantry cnc mill, so the cutter is sufficienty eposed to the work. You are talking very small travels &feed speeds on your Z axis so you could use low gearing of the Zaxis stepper drive to cope with the weight of the cutter head attachments &/or have counterbalance weights.
I think you SHOULD make a vacuum hold down table! They are so easy to make! I have found that particle board is porous enough to not require any hole drilling. With a good vacuum cleaner you can get at least 3psi of suck down pressure, and a vacuum chuck is so easy to build, especially since you can groove &surface it flat on the cnc mc itself. I just make a waffle/chest using thin boards on edge glued to a base board then surface machine the whole lot dead flat prior to glue (or screw on) of 16mm top particle board or thinner mdf. The same material should be used for the whole chuck to avoid distortion from varying humidity.
Anyway I have blazed on enough! It all sounds like an exciting project! I just love designing and inventing... have done lots of it all my life... 63 years old & still have fun dreaming & scheming & building, with loads of succesful projects under my belt. I always try to keep all options open in any design. You think of all sorts in the planning stage!

[dewme5]

Take this post as food for thought. Without seeing everything in detail, this is how I envision the system.

If you are going CNC, then Go CNC. Using the Z-axis, control the height of the cutter head. 1/16 is the max cut stated earlier (not sure if it which planner it was for) anyway, that max cut is what is acceptable with the feed rollers set the way they are. Drive the feed rollers as your Y-axis. Front and rear feed rollers will use gear reduction to up the power, and slow speed, allowing for a better and deeper cut. (remember, we are taking a manual use for everything planer, and changing it's purpose. 1/16th is the max, but that's set with the possibility of cutting a 20" wide piece of hickory or maple at the stock feed rate. Neither apply now, so why would the max cut depth?) Front and rear rollers will be driven at the same time, but off seperate motors to allow 2 cheaper motors to provide more power then one large more expensive motor - totally optional- however, both front and rear feed rollers need to be cnc operated.

Now, depending on actual work piece, while some may be accomplished in one pass, surely there will be a time when a minimum of 2 passes are required. While the optical sensor is accurate, how accurate is in question, and the profile for many reasons, needs to be repeatable. With the cnc control, the ski, once inserted, will not leave the planner until profiling is finished.
pass1
y feeds at a smooth slow rate, z adjusts to known distances vs feed speed to create partial profile, then z retracts to a safe distance, y returns to initial start point, and starts pass2. repeat until finished.

With this, the complexity of the cnc conversion is only increased by adding another axis. However, the software, drivers, power supply, wiring, etc. is taken care of when you put in just 1 axis, so adding just another drive motor is very simple. Allowing now, for complete computer control over the process, and 100% repeatability. Also, allowing now for replacement skis provided one takes damage two years after you switched to a new design, and quit producing the other.

[skibuilder]

Wow, you guys... those are some great ideas. VWSatOZ, you sound like you understand quite a lot about ski design/general engineering. I would really like to respond in detail to these latest posts, but the wife and I are heading out for 3 days of skiing and I have to finish packing all of the stuff into the car and hit the road. I'll get right back to this thread as soon as we return. Thanks again for the interest in my project and for taking your valuable time to share your thoughts and ideas. I can't wait to share pics of my cnc router with the cnczone folks. It's based on the Joe's 2006 model but has a 97 inch long X table, and is driven by a 1/2 inch ACME screw. I've added some do-dads that seem to have done the trick in preventing the dreaded whipping of the long drive screw. I don't have much to do to get the router finished up, but since there's finally enough snow around here to do some on-snow ski design testing, I won't be spending as much time in the shop for a while. I'm probably a month away from having finished router pics and video.

Happy Holidays,

Dana

[Rick Turner]

This could also be done easily on wood shaper using core profile templets and a vacuum fixture to hold the blank to the fixture. You could even power feed it through. I'd go with a spiral cutter head. You might also check on chain drive "automatic" shapers that would allow you to climb cut these parts. They use a sprocket drive and a chain built into the fixture for feed.

All in all, I think relatively conventional machinery would be quite cost effective on this.