SainSmart donated a PROVerXL 4030 to our team so we can get into CNC. It’s a lower cost ($1k) desktop CNC with a 400mmx300mm usable area that can be expanded to up to 1m^2. We’re looking forward to getting into it, but it only came last month and we as of yet don’t have build space so haven’t been able to do much. Our primary uses will be cutting plywood (for prototyping) and aluminum plate (1/8"-1/4") for the bot itself.
I did get it cutting 1/4" aluminum plate by adding a 1/4" end mill and taking a guess at the parameters. Only used it a couple of times in my basement so I haven’t figured out optimal parameters. It’s not terribly fast but did a decent job.
Assuming we do get build space and can compete this year, we need to get up to speed quickly. Has anyone used this on aluminum and has some parts/techniques to share? I understand we need to spend time learning it, but anything that gives us a head start would be appreciated.
@seanw I am not familiar with that specific machine. However, a lot of the advice you will find here and elsewhere regarding CNC routing will be relevant to just about any machine. Mostly, that will come down to:
Feeds and speeds - getting this right for your material and tooling is critical. It is a function of the tool, the material, and how much the machine can handle (how much it flexes, spindle power, etc.) Start with what the bit vendor recommends for chip load (or generic tables) and adjust based on your results.
Tooling - aluminum can be tricky stuff to cut well. Probably you will want a single flute endmill in the 4-5mm range for most work. Wood and plastic are more forgiving. Even if you end up using the same bit type, you probably should try to have “aluminum use only” bits that never touch plywood/MDF if you can manage it; the glues in those will dull bits and make them perform poorly in metal. Regardless, buy extra bits; you’re going to break some as you learn. There are some decent value ones out there, don’t spend $40 on a bit if you don’t need to. Try to standardize on one or a couple of sizes to minimize bit changes.
Workholding - there are various ways to hold your workpiece. We like to bolt down a sacrificial MDF spoilboard and drive short screws into it to hold the workpiece. Use brass screws so if you hit one, it’s less damaging than hitting steel. You can usually reflatten a spoilboard once or twice as it gets chewed up from use, but eventually it gets replaced.
CAM - hand in hand with feeds and speeds, learn to tell the software the details of how many passes at what depth, rough and finishing passes, etc. Details are specific to your CAM software. Since you’re already cutting I assume you’ve at least figured out the basics here.
I did look up your machine and it’s on the small side but plenty big enough for a lot of FRC parts. Size aside, the main limiting factor appears to be that it’s a pretty lightweight spindle, assuming you have the stock DC 24V 300W one and not the Dewalt router upgrade option. This will mostly constrain how much you can cut in a single pass.
Good luck with your new machine! Having CNC capability in-house can be a game changer.
Looking at the listing, and having used similar machines, the 4030 looks like it’ll be suitable for those materials. The frame looks beefy enough for aluminium, and the 300W spindle motor should do the trick given light cuts.
The spindle collet also looks like a real collet, a good sign as many budget CNC router manufacturers cheap out there. Try not to cheap out on endmills, single dollar Chinese cutters seem appealing but you’ll get much better life and finish out of a $50 on a few appropriately specialized cutters. If a cheap one catastrophically fails the cut, you may have to scrap the work which will cost more in the long run. Single flute cutters are a breeze on aluminium, similar for plastics.
Fusion 360 CAM is pretty good, would recommend it to finely control feeds, speeds and types of cuts in three dimensions. Fully featured, can’t ask for much more there specifically.
A basic plywood, plastic and aluminium cutting router is probably the beneficial tool in FRC for the $1200ish price range in my opinion. Hope you get alot out of it.
Thanks guys. I’ve been testing with a 3 flute 1/4" endmill but seems I should have gone with a 1 flute for better speed. Message received on quality vs cost!
We have been using Fusion in the past but are trying to move to OnShape because Fusion ties up our good developer laptop. Kiri-Moto has worked ok so far but I’m sure we’re missing out on a lot.
So far been very happy with this thing. My son and I have been playing with it in the basement on wood mostly, but some aluminum. I’ll update when we get to use it for robot stuff… thanks for the advice!
I did try making a PCB on it earlier but my design was too fine for the bit I used so it ended up taking away too much copper. With the built in Z-probing it was easily able to make a height map of my PCB and everything was even, at least. I’ll get back to it eventually.
Thanks for the pointers on the Facebook group and the Fusion Web. I didn’t realize Fusion could be run on the web, I’ll have to get the CAD people to look into that!
Last time I used Fusion on the web, it was very difficult to get any posted code out of the virtual machine it was running on. It’s been a couple years so maybe it’s different now. If I recall, there is Firefox installed to the virtual image that Fusion web runs on, but it was not easy to figure out how to launch it. I think in the Fusion options there is a browse button to set your post processor file path, and I think you can use that Windows explorer dialog to browse to Firefox, and use Alt-Tab to switch between them. Anyhow if you have a Windows or Mac computer, just instal Fusion rather than trying use the web version.
Native Fusion is a pig. Our team only has one computer that it can run on, which the developers also need. Few of our students have laptops to bring other than school issued laptops which aren’t that great and don’t allow software to be installed. So if we want to do CAD, it almost has to be on the web (which is why we have been toying with OnShape)