I don’t think it’s so much a scam as money-for-old-rope - it looks to be a repackaging of the open-source builds you can find all over the internet already (although it’s hard to say, because the promo video is so glossy and superficial).
I looked at some reviews of this video tutorial/documents and it does look like some effort has been put into the production, but really, I don’t think there’s anything in there that you couldn’t find for yourself on instructables, youtube and various maker blogs.
So you’re paying $39 to have someone else do the research for you. Not too bad of a deal.
Rods are the way to go.
There’s no way that wheels in channels will be free of play (unless you use pairs to grab a rod, in which case you might as well use a plastic slider on a rod).
But, you can buy low-play linear bearings really cheaply. Why would you want to make such a critical part yourself?
Yeah, the only concern I have with rods is potential sagging in the middle of the longest horizontal axis - channels can be very resiliently supported. Rails can be supported in a way that doesn’t interfere with pinch-wheel pairs running on them. Rods just need to be engineered thick enough that they won’t significantly bend in the middle under the weight of the gantry, motors and spindle.
Assuming a 1 metre Y axis supporting a gantry that is, say 25cm broad at the base (so maybe 70cm of Y axis travel), how thick would you make the rods?
Here an interesting kickstarter – it doesn’t use rails at all:
https://www.kickstarter.com/projects/1830738289/maslow-cnc-a-500-open-source-4-by-8-foot-cnc-machi?ref=category_newest
Brian
Yeah, I spotted that one in my searches - really interesting, but the project appears to have gone full circle through funding to rollout and now the product itself is no longer available.
They cite a surprising level of accuracy for this machine (0.4mm IIRC), which I found a bit hard to believe, given that the cords supporting the thing must have some elasticity to them.
If you are worried about sagging, you can get supported rods, with split bearings.
Here’s a reasonably priced set, including ball screws:
That’s reasonably priced for what they are (and they would be perfect), but way out of range for my project.
Well, I just took the plunge - I actually just ordered one of these - because that’s a very good price.
It’ll need an old PC with a parallel port to run it, but I’m treating this as entry-level - my cunning plan is:
[ul]
[li]Treat this as a very definitely ‘entry level’ device[/li][li]Use it to acquire the necessary skills to run the thing[/li][li]Study the design and its good/bad parts[/li]li build my own larger device at a later date - perhaps by cannibalising this unit and rebuilding it on longer tracks etc, and/or using this unit to manufacture some of the parts.[/li][/ul]
You know you can get USB-Parallel cables for real cheap, don’t you?
USB->Parallel cables don’t usually work for these type of things. Most of the parallel port based CNC controllers are going directly to the hardware and bit-banging the pins to set stepper direction and trigger the steps. It’s very important for the timing to be accurate, and USB-Parallel converters mess that up. Also, the CNC hardware usually expects 5 volt TTL where as the USB cables (and some bulit in laptop parallel ports) are 3v.
I have plans from a couple of different sources, and have built a HobbyCNC controller board, but I have yet to actually build out a whole CNC system. It’s kind of overwhelming, and I’m afraid of wasting time and money on things that turn out to not work at all. At this point I’m very tempted to get something like an Xcarve because my time is worth more to me than the price difference.
I’m late to the party, naturally, but you can get parallel port cards that plug into PCI Express slots. Startech do one here.
For a hobbyyist user with a dedicated machine who uses the machine only for CNC tasks and accessing a handful of CNC related sites like thingiverse, your concerns are approaching tinfoil hattery. Machines with limited if any browsing very rarely have problems. We have about a dozen internet connected machines driving various CNC/3d printer/laser cutters. Never been a meaningful problem
Yeah, many of the parallel port controllers I have seen on sale explicitly state that they need a computer with a native parallel port, and not a laptop.
Here’s where I’m confused though - they specify parallel port and Windows XP, but XP has the Hardware Abstraction Layer - I thought direct hardware access like that went away with Windows 98.
If you transition from being a home hobbyist to being a paid business you’ll want to be more careful. Some bad guy takes your machine down when you’re a business, it’s your bottom line that suffers. Another concern is your networked machine being made part of a botnet - unless you’re monitoring your network traffic you might never know.
My point is risks are there, and must be approached as thoughtfully as any other engineering decision. I’ll need to write up a more detailed approach in my copious free time.
How will monitoring your network traffic help you to detect nonexistent attacks on machines that are set up as standalone, off your network?
Huh. The Amazon seller cancelled the transaction at the last minute. I thought that price was a little too good to be true. Back to square 1.
Turns out it was a scam. The first time I’ve actually encountered one on Amazon.
The machines are back up and listed now, at bargain prices, but they are all marketplace items, and they all have a comment somewhere in the ‘condition’ notes saying that buyers should contact [some email address that has been munged to avoid detection by Amazon’s system] before buying.
I emailed the address, out of curiosity, and got this response:
Scam. They will generate a fake Amazon invoice and take payment through some method that is not Amazon, then disappear with the money.
What goes from software bit twiddling through the OS HAL to a physical parallel port is close enough to work.
What goes from software bit twiddling through a virtual parallel interface to a USB interface to the USB HAL to a physical USB to a cable with an embedded microcontroller that tries to generate a parallel signal from the USB data feed, all while costing less than 2 cents apiece with fine Chinese driver software, well … not so much.
Sorry to revive an older thread, but wanted to hear how things are going.
I’m getting almost ready to make this jump too. I’m sure I can make the base, x/y/z axis parts, it’s the motors and software that are intimidating.
Planning on using Maytec aluminum extrusion with slides as the primary x-axis bearing, with an acme screw to drive it. Using that extrusion because we build museum exhibits and have used quite a bit in the past, and there might be enough kicking around in the shop as scrap.
Anyone else watched Frank Makes a cnc? Helpful, but I know how to do what he shows there.
And now 20 minutes later, after getting lost in youtube videos…