Have you got illustrator? Open in illustrator and export as dxf?

 

This any good?

https://3dwarehouse.sketchup.com/3dw/download?mid=5cbccb74c904efe49d99f2f186e46f1c&rtyp=k2&fn=Fender Signature Stratocaster&ctyp=other&subjectClass=entity

 

Sorry every sketchup file I get is a complete headache. I've tried all the dxf settings to export better quality dxfs. Broken vectors and dimensions that don't stick and float +/- >0.1mm in all of the settings. Vectors on top of each other. If it's an export problem then the tens of drawings I've had to redraft from various clients must all have the same wrong setting.

I use both solid works 2014 and inventor 2016 and both experience these issues.

Don't get me wrong it may be okay using the inbuilt cam, not tried it so I can't comment but its not there in terms of use in a professional sense.

 

Maybe you're not there as a professional yet? Tons of pro's out there using SU and making very sophisticated designs. I use SU now for years and have had non of the issues you're talking about.

 

sorry praux forums are places to voice experiences, this is my experience with the dxf's i receive made in su aswel as the fixes I've tried. Yes su can make fairly sophisticated designs. but thats not what im talking about here and you wouldn't have experienced any of these problems unless you use other 3d cad packages. Once you have created a few hundred tools path and you are doing it to make money you will find this out, You still have a good 6 months of getting used to your ox before you will find that you need your software to be efficient at sharing files and that yes su will get you started but it's not going to do everything you will need it to do.

 

The only catch to that line of thinking is that most people see their local data evaporate long before the cloud. Thus for the average user, having a copy in the cloud is often best ... as much as I hate saying it (I have similar thoughts - Google, for example, does indexing like all others and this means their computer do read your private / confidential information regardless of what they say).

One needs to have backupS (more than one), especially of their critical data, no matter where or how it is stored. The cloud becomes an easy backup by its (relative) persistance compared to what the average user might do for his (her) local backups. It is also remote - potentially half way around the world, unless you happen to live next door to one of the secret data centre. If the mega meteorite ever strikes the planet, no one will care about any of the backups.

The coffee cup is more likely to spill at home then in a bank of racks hosting a cloud.

The DVD might outlive most users and maybe some cloud providers, but will anyone be able to read the DVD 20 years from now? How many remember the diskettes, ZIP drives, ... ?

Even with all the care in the world, hard drives will fail, even the SSD devices will fail with time. The ones in our computers before those in racks used by data centers with controlled environments.

To have a paper copy ... well time will also get to that. So have as many (or few) backups as you can live with or afford. However, be certain to refresh your backups even if the content is "static", be for updating the technology of the backup or just have a fresh copy and see the content is still readable by at least you.

In my 30 or so years in the computer industry, the weakest link has always been the people, backups all too often a forgotten aspects ... until needed. This reminds me, I better go through my backups !

 

I have autodesk a360 (free) syncing my docs up along as well as one drive, then I have a backup drive that syncs also. Just now trying to find out how to hide them in windows 10 as I have multiple folders in my explorer list which is annoying, have to be careful not to save to any other than the original folder.

 

In my first attempts to use SketchUcam I was getting quite frustrated ... even more than with SketchUp itself ...

What I have learned about SketchUcam while working on my first sign (see post under myOX link) is that one should breakdown a seemingly complex task into smaller tasks, just as seen in the "seatknob" video. I should of thought of this, being an aging programmer ! Remember the "top down approach" ?

SketchUcam, when taking this divide and conquer approach, can be very good and easy to use. However, it will require you do some prep work, as seen in the "seatknob" video. This actually helps in documenting the job ...

In addition, by breaking down a job one can also do tool changes. The roughing in the video could be done with larger end mill. In this particular example, the time needed to change the tool is probably more than what could be saved by roughing with the bigger end mill.

All this to say I am starting to love SketchUcam as it allows me to have more control. Of course, it won't allow me to do everything I can imagine needing, but it certainly will do a lot.

David (the swarfer) was of great help when I ran into 'issues' with SketchUcam, all where really me not having thought my project through, much less worked with the tool. By breaking the job into three parts, it went from over 10 hours with me scratching my head for days trying to figure out why SketchUcam was not behaving, to only 20 mins on the same CNC machine. The needed prep work took only minutes : take a copy of model, add reference point to keep things aligned within the work area and pick elements to be done at each step to then do merging of the resulting gcode files to run as a single job.

Of course, with enough money, there are CAD and CAM solutions which will do this optimization. My budget is limited, so I will continue my learning curve with SketchUp and SketchUcam (and MeshCAM since I have it)...

 

In your enhancements, have you thought of doing/allowing pockets with some form of spiral patterns ? Pocket circles, for example, from the edge in or centre out ? Maybe figure a way around SU offset issues by pocketing a shape inward while always staying within bound of previous trace ? If done as a continuous trace, it would stop once bounds does not allow any more material removal. Complex shapes would still require 'zoning'.

Of course, all much easier said then done ...

 

Hsm is your answer buddy.

 

Except HSM is for SolidWorks and the latter is not for my budget, unless I twist my status at an educational site hard enough to fall in the 'student/staff' crack to get a limited use (not capabilities) license for only 150$US. But it doesn't include HSM anyways ... Or am I way off in left field ? It wouldn't be the first or last time.

 

Its Free!! i pay for inventor pro but actually talking to the rep today and tried haggling on the price, which he wouldn't budge but asked me if i needed 5 axis if not and i quote " you could just use the free student version".

I have no reason to lie about this, its free and has been for a while. and yes it does have 3axis hsm.

http://www.autodesk.com/education/free-software/fusion-360

 

Sorry serge may have brought up a couple of conversations with other people back there.

The free license is for 3 years.. but it doesn't take a genius to work out a way round that. possibly asking your partner who so happens to be learning it too

hsm solidworks, inventor, fusion 360 is all the same engine.. there are varying levels of usability.. fusion 360 is the simplified version but works very well and performs all the same types of operations only with less variables

It requires no evidence that you are a student, You could also be a mentor.

 

http://www.autodesk.com/products/fusion-360/overview

Free for students, enthusiasts, hobbyists, and startups
Full use of Fusion 360 for as long as you need it

• Start by downloading the 30-day trial
• Once you’re in, simply register for free use

 

Including 'startups' ? It sounds like quite a few people could take advantage of the "free use" offer. Talk about a lost lead to get people into the higher end versions. I better check this out, eh ? Fusion 360 sounds like it can do plenty, especially for a person like me ... If only it could rain this weekend.

 

Yeah I suppose it's in their interest to get more people involved in it to sell the high end packages to later on.

 

Indeed, the math can get heavy ...

The 1.3b spiral pocketing of circles will be great. The fewer time a tool needs to move in the air, the more we save in run time.

Will it handle a circular ring as well ? There is no 'center' to start from ...

Could you make the 1.3b circle pocketing available for 'all' shapes and let the user cancel out or add line segments to simplify the shape as done with the current pocketing routine ? I suspect it would work fine with ovals (a stretched circle) and other simpler shapes, right ? It would probably work with anything the SU offset routine can handle. It would be one more step forward with spiral pocketing ...

 

suppose that's the difrence between hsm and su. Su being more restricted to basic gcode moves.

 

Indeed, it looks like SketchUp and SketchUcam, although with a lot more control and not having to copy the model for each group of milling steps ...

I watched a video where the person works with a model of a gas cap. He pretty well goes through defining each optimum milling steps for shaping the end product, just like it would be done with SU and SketchUcam. It would probably be much more work with SU since there is less control and one has to copy the model for each step and then join the g-code.

In any case, it's still not the CAD/CAM figuring out the work. MeshCAM takes the approach of going back and forth along X and Y axis until all extra material is removed. Great for really complex work, like carving a photo. Not the thing for relatively simpler optimizable milling/carving. No overhangs allowed using a 3 axis machine, unless it also one can do multiple faces/facets (do one, rotate however many degrees needed, do the next).

So far, SU/SUC are good enough for my needs. But I'll sneak a peek at Fusion360 and others as I progress and leave my confort zone. MeshCAM will be used as well, where more appropriate or when I'm too lazy to divide a job into a series of steps. Like they say, it can be 90% preparation, 10% execution.

 

Got it. There will be some interesting effects amongs those 64+ combinations. It will take some sacreficial waste material to play with them ... better yet, lots of OpenSCAM simulations.

Oh I know ... without the suggested tool, I used SU's offset tool to do what you suggest with my first inlay work :



It was ... interesting, more than fun ... especially as I neared the end and realised I was doing a 50% diameter offset instead of the recommended less than 30% or more than 70% tool diameter. Anyways, there was tons touch ups as SU offset went bonkers in tight spaces, sharp turns and corners. But it worked.

Indeed, I had the advantage of seeing ... the brain is amazing, even my lonesome grey cell ! Worse yet, SU seems to give those points in a random order from what I've been able to gather thus far. Making the job of a addon tool that much harder ...

Please do, especially with the latest release of SketchUcam.

 

I would not point to the basic g-code moves available, but how they are put together to get to the end result. Like DNA, there are only a few 'basic code' but what a diversity we get when a bunch of them are strong together ...

I can take a model and feed it through MeshCAM to get a back and forth X and Y series of tool path in seconds or take the time for describing the components for a (near) perfectly optimized tool path with SU/SketchUcam (or a fancy tool like Solidworks/HSM) to get to the same end results. The g-code generated use the same basic moves. However, the first option makes my work real easy but the machine's a long boring process. The 2nd flips things around to make my work potentially real long (but interesting / challenging) while the machine flies through the milling in no time to do the same part.

There will always be more than one path to reach a destination. Even if all roads lead to Rome, I have never seen it... in person. Some like off-roading, others like a chauffeur'ed drive. I like side roads, lots to discover on the way.

 

Yep I think that goes for all cam in that it most definitely a case of stringing operations together, it's like you say it really comes down to how quickly you can do it.

The big difrence in cam software is some are capable of combining multiple types of operations to happen all at the same time, on others to do this the gcode would require complicated splicing manually.

So have I got this correct that you find meshcam quicker for roughing over hsm? I pretty sure that hsm can do it faster as it supports morphed spirals.

If you really want to rough a part quick the hsm 3d apadptive operation with morphed spiralling 95% step over both directions with the correct settings seems to me that it would be impossible to better as the cutter maintains the set engagement for as long as possible, it combines 2d and 3d roughing operations and using morphed spiralling reduces the number of deaccellarations to a minimum and takes roughly 40 seconds to create the code as you need not select any of the features, all automatic.

All that said i never actually use this and prefer the aproach you take in that I like to set my operations manually.

Just to add I wouldnt recommend sharing the process between two difrent cam programs. Your almost guaranteed to get an error in the origin or part position between the two.

Edit, I just re read your post I may have read it round the wrong way, not sure.

 

Sorry, I tend to ramble on at times one thought leading into an other and loosing track of the initial question or subject. But, hey, no one is perfect, right ?

I was just saying that some tools are quicker (more appropriate) for some tasks. It's rare to have a tool good for all tasks.

I like the sounds of the morphed spiralling with 95% step over in both directions. Maybe just one click in HSM, but not necessarely impossible with other tools ... just loads of prep work to come close to it. Add the use of biggest end mill for the job at hand and roughing could take no time at all.

I would love to see an affordable CAM solution which could take into account the dimensions of the raw block. In other words, not assume it is a cube - parallelepiped to be exact, since it could have bulk of extra material removed using more effective means... or be any other shape (ex: carve onto half sphere). This without carving air to reach the actual material, thus saving a lot more time.

There I go, off on a tangent again

 

H

Not at all, it most likely makes clear sense.

Yeah with you there, for instance hsm is not efficient for 2d with multiple nested parts as say vectric software.

I suppose hsm can do that in a way, if you were to model the shape of the stock and then create a stock setup around that shape, ie that will fit the setup stock box around the shape, the use a 3d apadptive operation to make the stock the same shape as your model or physical stock, doesn't matter about how the cam does it as your not going to post process it, you can then use rest machining with setting "from previous operations" for all the rest of the operations this will then assume the remaining stock is that of the physical stock shape.

Obviously you will need to delete the model of the physical stock to reveal the actual model you want to mill. So the operation will have an x on it in the browser but in my experience it seems that the toolpath remains and the rest machining still accounts for it.

Hope that made sense. Like a virtual process to take the stock box to the physical stock shape.

Though a way to add a model to represent the stock without all of this would be really handy.