[jbaz]

Coming from 3d printing (additive) the process from design to g-code is very simple. Create part in Solidworks, export STL file, slice into G-code and press print. I have been running manuals machines for a while but now that I have a couple hobby/trainer grade CNC machines properly running I'm having trouble with the G-code generation for 3-axis milling.

In my mind I was hoping it was going to be just as simple as additive printing with the addition of a tool table that contained my available tooling. So I thought it would just take my Solidworks part and the dimensions of the base stock and just generate the optimal g-code from the available tooling to cut the part as accurately as possible, with the option to the tweak the settings to for precision vs speed. That is what I was hoping for, but what I have demo'd (admittedly not many yet, but I have read up on more)... they don't work like that at all.

For background, I will be doing about 80%+ 2.5D work and likely no more than 20% 3D (multiple operations, manual flips). The bulk of the parts can be finished with no more than 3 endmills, usually just one or two. These aren't terribly complex parts, but because I am familiar with Solidworks and can quickly sketch a part that will be 2.5D I do it in there. I haven't touched any other CAD or 2D drawing packages in a long time, couldn't even tell you where to start on anything else.

None of this CNC is for production or even with a profit motive, and not just because of the limitation of the machines either, it is just for hobby. So even a few hundred bucks is it a little overboard for this stage of the game, let alone a couple grand. I'm hoping perhaps there is a good open-source offering or a hobbyist centered package out there for a reasonable price that really simplifies the process where you're doing a lot of one-off stuff and are looking for speed (from model to g-code) instead of precision or optimization.

I have tried the demo versions of CamBam and SheetCam and I'm sure once one is experienced with either they can be very efficient and effective but this approach is quite different than I am used to and actually felt like I was working with very dated software. Maybe the real magic is just supposed to be under the hood, I don't know.

I guess I am looking for something more intuitive and robust from the standpoint of the software doing as much of the lifting as possible with the least amount of input using a Part or other 3d model file to produce both 2.5D and 3D (manual flips). Anything out there like that, and if so, ideally something still in the price range for a hobbyist?

I was just about to download HSMExpress. I don't know much yet except it is free and works with SW so perhaps that could be the ticket. It is just hard to spend enough time giving every demo a fair shake when there are many options because it can take a lot of overhead just to figure it out. I also heard Deskproto might be the way to go. I think someone mentioned it has intutive features in 3D where if it has a non-flat surface(s) it can auto create landings on each edge of the part so you can secure it squarely in the vise and then cuts them off at a later operation... that is the kind of automation I was hoping to get.

[jbaz]

Update... so maybe my idea of the ideal CAM package I wanted was asking a little much, but I got a chance to try very DeskProto for a brief period and so far that seems more like what I have been looking for because it can take a STL or other 3d model and seems like the wizard is very simple to get g-code output for both 2d and 3d (manual flipping the work piece). I also checked out HMSExpress and it seemed very robust for professionals but reminded me of the other CAMs that seemed like they had more overhead to get g-code out. It definitely was going to require watching some training videos on it I think.

So now that I have a point of reference, anyone know of other packages that have a similar approach as DeskProto? For now I'm just looking for a way to get my existing models cut to see some progress as opposed to having optimal g-code or precision parts cut to keep the progress rolling.

Thanks!

[ger21]

MeshCAM

[jbaz]

Wow, thanks! I don't know how I missed it because it was on my list to review (read the description, check out screenshots, etc). I wrote down all of the CAM names from the CAM subforums that had a decent amount of posts in them figuring these were the most popular and likely the most worth investigating. From the website it sounds right up my alley and a similar low-overhead approach to getting g-code from already created parts.

I can respect the need for the advanced options present in a lot of the other options out there when it comes to creating very precise work on capable machines doing production work, but this little Sherline is far from that along with my abilities so these seem like the best that suite my needs right now. I almost don't want to demo MechCam yet because it has just a 14-day trial and that might be tough to learn both well enough to make a decision in such a short time when I only have a couple hours here and there.

I am still open to more suggestions though to anything similar to Deskproto and MeshCam. Even at a few hundred bucks the software is a little pricey for purely once and a while tinkering/personal education so I definitely want to pick the right one the first time.

[christinaB]

Well, what I can say is be yourself. I know, you're creative enough to think of the best. MeshCam is a good one but you can make it better.

[ger21]

If you contact the Meshcam developer (Robert), he'll extend the trial time if you need it.

[jbaz]

Ok, thanks for the tip about extending the evaluation. I am still trying to get through DeskProto which is remarkably simple yet seems fairly capable for the type of machine and work I am doing. I think MeshCam might be a level up from it in terms of options which might be helpful (hard to say without using it yet - only reading).

I was finally able to get FreeMill working. I didn't know you had to start VisualCAD as administrator to get it to work right because it would just go straight to VisualCAD without any option for FreeMill running it normally (at least on Win 8.1). Anyone doing very simple or infrequent work it seems like a good g-code generator for the price (FREE) because it takes 3d files and creates the tool path. So far that is the only option I have found that is free and can take 3d parts you've already modeled in other software. If you're doing hobby work on a small hobby machine without much much of a budget for software this is the route I would go.

Question... why does there seem to be such a small offering in the way of lathe CAMs? Is it just because a lot of people end up doing lathe style work on their 4 axis mills, or because doing manual work on a lathe is much easier compared to mill so there aren't a lot of CNC lathes creating the demand?

[awerby]

I think DeskProto has more options (such as 4th axis machining) than MeshCAM, but both are a step up from Freemill, although they do cost a little money. All these programs basically work the same way: they start with a mesh file (STL is the most commonly used format) and cut the whole up-facing surface, although you can define the areas that get cut. Complexity of the surface isn't a problem; the choice of which ball-end tool is used determines the level of detail that will be attained. While this is analogous to 3D printing, and certainly simplifies things, it's not the way machinists usually go about it. Machined parts are generally cut using "2.5D" procedures - essentially defining flat 2D areas like pockets or holes with 2d shapes and cutting them with flat endmills to certain depths (hence the .5 D). Some more advanced CAM programs, like VisualMill (from Mecsoft, the makers of Freemill), combine mesh machining and 2.5D features so that parts can be made either way, or both ways, if they have both kinds of features. The choice of program really depends on what types of parts you want to do.

As to your question above, I think it's the latter - there just aren't that many hobbyists using CNC lathes. While there are some lathe-type parts that can be made, at least partially, on a 4-axis mill, it's certainly not a substitute for the lathe. But most hobbyists still use manual lathes, even if they have CNC mills. There certainly are pro-level programs for lathe code generation - Mecsoft has one called VisualTurn - but there's not much available geared to the casual or hobby user. But programs like that do exist - try checking the newly-revised "Links" section of my site, under CAD/CAM software, to find a few of them.

Andrew Werby
www.computersculpture.com

[Reliefs3D]

Have you thought about ArtCAM Express?
It does a pretty decent job in 3D milling but 2D is simply outstanding.
check it out ArtCAM - YouTube

[jbaz]

Andrew, thanks for that detailed response. So far DeskProto has done a good job but there are some nuances to it that I feel could make it better when doing more than one-sided milling. I am going to give MeshCam a try soon to see the difference first hand.

Reliefs, I did look into ArtCam but it appeared to only do single sided 2.5D milling. Although most of the parts I will create will be 2.5d there are quite a few that will need full 3d (manually flipping the work piece for 2nd side). I plan on using a demo of it though to confirm and get an idea of the features by other platforms so I know what is possible.

[mmoe]

My guess is that Meshcam is going to be the best option in what I gather your budget is.

As for lathe work, another factor for why hobbyists don't generally need CAM for lathes is that not only do they have manual lathes, but they really don't have a lot of need to upgrade from manual to CNC for lathe work. Most of the work with a lathe is setting it up, so until you get to the point of having a large turret lathe, a CNC lathe really doesn't save a person a whole lot of time vs. just doing it manually. There have been many projects where the setup portion of the lathe operations has been 80% or more of the time spend making a part, so saving even 50% of my time on the last 20% of the effort just isn't a big enough savings to make it worthwhile, especially given the expense difference. At least that's the way I see it.

[Tkamsker]

I use deskproto and bobcad so i can do lathe aß well

Gesendet von meinem SM-N9005 mit Tapatalk

[erikamaeB]

I have a hard time with it. Think of the manual then.

[Tkamsker]

I think of an small toolchanger for my lathe Which connects to digital x and z calipers so i think this Makel Sense in Manual and cnc work just a Button and it jumps to Desired Tool and Tool will be good aligned ...

Gesendet von meinem SM-N9005 mit Tapatalk