New Shark Pro HD

[jeb2cav]

Here's Greg's project file. Not sure why the crv won't load - but it won't.

Some thoughts I have by looking at the project - Can you look at the tool database and see what your pass depth is for this tool? It looks to me like the VCarve default setting for this 90D vbit is 0.5”. I think that is too aggressive a depth. From my reading, a recommended pass depth for an end mill cutter is ½ the tool diameter. The tool in your project is a 90D vbit that is 0.5” wide. So this tool is currently trying to cut using a pass depth that is the same as the width of the tool. Yes, it’s not an end mill, but I don’t think you’ll get good results with this pass depth.

Last week I ran a VCarve in black walnut using a 60D vbit that is 0.6” wide using a pass depth of 0.2 at a plunge of 15 and feed rate of 60 ipm, and it was fine on my ProPlus. I have found that I spend less time sanding and get crisp edges with a lower plunge rate, and by fooling with the pass depth and feed rate of course. I also use 18-24% for vcarving clearance pass. Again, these might not be 'the best', but provide good results for me.

Precise bits has a pretty good guide on how to determine feed rates. From reading this and also some other things available from Onsrud, Vectric Forum, etc – the layman’s advice is to start conservative, do some test cuts, and work your way up to the max. Often times having what may seem like a slow plunge rate, but still having a reasonable feed rate and the sweet spot depth of pass gives you a really good result, and at only a little more time on the table than you might have expected going in cold.

[REG]

When I look at the file it has a number of toolpaths that seem to duplicate other toolpaths. I checked vectors/nodes, those were in proper order. I made a new toolpath using a V90 degree 0.50. Ran the toolpath and the project came out perfectly with a max depth of 0.125 and just under 3 minutes to carve with 50 ipm feed rate.

Bobby

[kruizer]

When I look at the file it has a number of toolpaths that seem to duplicate other toolpaths. I checked vectors/nodes, those were in proper order. I made a new toolpath using a V90 degree 0.50. Ran the toolpath and the project came out perfectly with a max depth of 0.125 and just under 3 minutes to carve with 50 ipm feed rate.

Bobby

Bobby,
I ran your updated file on my machine and came up with good results. What did you change in the tool path?
kruizer

[REG]

kruizer, I eliminated all of your toolpaths from your file that Joe posted and made one toolpath from the vectors you have in your project. If you want a side by side comparison, VCP will allow you to open multiple windows of the program. So open your project file in one window and then open the one I posted in another window, adjust the size of the windows of each to have them side by side.

Hope that helps.

Bobby

[DickWaldron]

I'm new to all this, but are ther programs out there that will allow you to download a picture and get the shark to carve out the picture? Dick (considering buying one).

[fison]

If you're wanting the photographic image photoVcarve will do that. If you're wanting a 3D representation of the photographc that gets a little more complicated and not something I have any experience with. If you're wanting to carve line-drawings, etc. VCarve can do that will a little effort.

[kruizer]

When I look at the file it has a number of toolpaths that seem to duplicate other toolpaths. I checked vectors/nodes, those were in proper order. I made a new toolpath using a V90 degree 0.50. Ran the toolpath and the project came out perfectly with a max depth of 0.125 and just under 3 minutes to carve with 50 ipm feed rate.

Bobby

Bobby
Many thanks for the help. I ran your modified file and it turned out perfect. I have learned a lot about using VCarve to do lettering through you all. I found a piece of the kids tree that I could use to replace the part that was lost and all is good again. Happy Holidays to all.
kruizer
PS My HD still does not make a round circle. Still egg shaped by about .030"

[REG]

Bobby
Many thanks for the help. I ran your modified file and it turned out perfect. I have learned a lot about using VCarve to do lettering through you all. I found a piece of the kids tree that I could use to replace the part that was lost and all is good again. Happy Holidays to all.
kruizer
PS My HD still does not make a round circle. Still egg shaped by about .030"

kruizer, I made a simple circle file. I have zipped the file and the toolpath just as I did on your other file. On a 12" X 12" this will carve a 10.5" circle that will go about .125" deep. Use a 1/2" V90 bit. You can safely clamp the corners to carve this circle so that you can be certain the material is tight to the table. Once you run the file, if the circle is perfect as shown in VCP preview, then your Shark is fine and the problem may lie with your selection of the Post Processor (when selecting the toolpath and save, make certain the Post Processor is set to CNCShark-USB Arcs (inch).

If the circle is not a circle then the problem is with either the Y or X axis coupler and/or ball screw nut. Most likely the problem is with the coupler. As mentioned before, if there is "play" in the movement when you push linear to Y axis or X axis you have a loose coupling. If the couplings "were" loose and you tightened them already, you may not have taken out the unload play of the axis screw; meaning that you tightened the coupler but you did not pickup the clearance play axis screw. With the couple set screw loose, you should pull the axis screw into the coupler.
As a side note, I find that when the Shark controller is on, the stepper motors are energized locking the spindle of the motor (rotor) to the center. When the controller is off the motor spindle will "float" moving within the clearance between the motor end bearings. I set my coupling with the stepper motors energized.
Also, when setting the coupling, make certain both end set screws are tight and are afixed to the flats of the shafts they couple. That may seem like common sense but it is easy to be on the edge of the machined flat area. The set screw can roll off if it is not at the very center of the flat area.

I would be interested in knowing if either of the suggestions were a solution to your problem.


Bobby

[rungemach]

You may want to repeat the cutting of the circle using a flat sided bit if you are have any trouble measuring to the bottom of the V cut for the circles dimension.

You can't go off of the top surface measurement as that will change if you have any Z fluctuation over the surface of the table, and you don't want table flatness distorting the measurement if the z height is not constant..

Mark your part so you know later which direction is 12 o'clock when you start measuring diameters.

Bobby is right, eliminate the backlash issue with a simple test , then you can go on to other reasons.

True backlash should not change much with the size of the circle being cut, you can go smaller depending on what type of measuring tools you have available. If it is a backlash issue, a 1" circle would be about the same amount "off" as a 10" circle, as the backlash amount stays pretty constant.
If it is an axis calibration issue of steps per inch, it will compound with size. If you have an inexpensive 6" caliper, 1 inch and 5 inch are good test sizes.

For a Shark, the lead screw nut should be under .002 backlash, so your .030 is quite a bit larger and something may be loose. The design of the shark requires the shaft couplers and motor bearings to resist both rotation and axial forces that "work" on the coupler to try and vibrate it loose. The flat in the shafts help the rotation resistance quite a bit, but do not do much for axial ( push / pull) forces. It puts more forces on the couplers than a bearing supported lead-screw design, and hence there are issues in keeping the couplers tight. ( a recurring theme)..

Hope this helps.

Bob

[kruizer]

Well Fellas,
A lot of the problems were to a large extent solved. Jim at Next Wave had me drop the gantry down on my machine to the top most holes and the chatter has mostly disappeared, the out of round condition is not as bad (maybe .008-.010), the step condition has gone away. I also raised the router motor up so only 1/2" is showing below the clamp. This means the dust shoe cannot be used. Just as well as the dust shoe was holding the Z axis up about .020" and distorting my work a bit. There is still a hesitation at the direction change at the 90 degree point and 270 degree point. We don't know what that is yet but sure it will get worked out. That may be what is causing the out of round condition. Had to take the Porter Cable 890 motor back to Rockler as a bearing started getting noisy after about 20 hrs. They were very good about the exchange. So life is getting better.
kruizer