Has anyone ever used or gotten feedback from anyone using this controller board?

Has anyone ever used or gotten feedback from anyone using this controller board?

The better way is to use a controller whitch is connected via Ethernet.

George check out the Buildbotics controller. I just got mine in and it is awesome for the price.

I use a plain old PC. I picked up the machine I’m using at a senior center sale for $5. Can’t beat it. Though not every desktop is really suitable for real time control. But with a full PC I have nice file handling features baked into the system. None of this I have to use a SD card crap. I use free motion control software too. LinuxCNC works for me.

@Jared_Cochran I’ve been curious about that. I emailed one of the head guys about the kickstarter couple months back, but I wanted to get some feedback from customers before I purchased it. I’ve been watching that for a while. Cottonfield or Crowefield was his name. Anyway, it’s great to hear about that. I’d much rather go that route, plus they have a prebuilt enclosure. What type of machine do you have it on?

@George_Allen what do you need all of that crap for? All you need is a $5 break out board. heh when I Google for break out boards mine shows up. https://www.instructables.com/id/Parallel-Port-Break-Out-Board-BOB/ I’m famous! But they make the same thing commercially. You want one with buffer driver ICs on it. I used 74245 AHCT chips. They have nice current output. 24 ma a gate. It is impossible to get a spec for parallel ports. Believe me, I tried. The best I could come up with is you’re guaranteed maybe 3 ma? I forget what I set my BOB to switch at but it is less than a milliamp. Ohm and his damned Laws. I know I can get 24 ma out of it too. An optocoupler on a motor drive might draw up to 20 ma but 12-15 is more average. That kind of draw could smoke a parallel port.

That’s the secret of electronics BTW. Do not exceed the current capacity of devices. OK maybe there’s a bit more to it, but if you just stick with that stuff won’t blow up on you at least. Which is definitely half the battle. The other secret is buy parts in bulk. That way when you do blow stuff up you have spares.

I bought one, just havnt used it yet.

George Allen I am using mine on a knee mill retro fit. I talked with Doug Clifford about what I am doing with it and they provide excellent customer service. He actually just called me a few minutes ago about using a stand alone monitor and keyboard on the internal raspberry pi. I plan on converting my laser cutter and eventually my plasma table.

No, I didn’t! It looks really great though I am not so sure why an ARM processor is needed because all processing (at list with MACH) is made by the computer. Why I would one 125USD when a few bucks chinese breakout board do the same. At list for me it’s good enough. Maybe if it could process USB signals… but it doesn’t seem to me to do such thing…

@Paul_Frederick I actually have a BOB.

@George_Allen then you’re set. Because BOB is all you need. Well, once you get PC timings worked out. Which can be non-trivial. That’s the inherent difference between Mach and LinuxCNC. The LinuxCNC team managed to do it right.

@Peter_Spiess “The better way is to use a controller whitch is connected via Ethernet.”

Why?

I am not saying you aren’t right, but just saying ‘the better way is’ without explaining why its better is questionable.

Ethernet better than direct parallel connection? Better than a dedicated controller board? I’d be surprised if that’s the case.

@Mike_Thornbury There are many points against parallel connection. At first the limited number of inputs and outputs. One parallel port have 5 inputs and 12 outputs. Speed plays a role as well. Ethernet gives me a rate about 400 kHz. If you use the parallel port, you will use a special Linux kernel or an old Windows XP. In this case there are no problems with ethernet. USB always have dropouts. I destroyed a workpiece because the milling machine simply moved on, although I had already released a button (handwheel connected via USB).
Since I connected the machine via Ethernet, everything works perfectly. That’s not just my opinion, many friends agree with me.

@Peter_Spiess , i very much doubt… Maybe if you have a very sofisticated ‘BOB’ (in fact a processing mainboard) capable to process itself the CNC code sent from the computer. Such way it would pe possible. But I am pretty sure such a ‘board’ would be hundreds of USD not a fiew bucks… In regular case, the processing is PC’s job (windows or linux, diesn’t matter) which will send necessary signals to the axis drivers. And this needs something what a network doesn’t have: enough SPEED.

@Alex_Paverman No Alex, you are totaly wrong! My Ethernet Motion Controller costs about 135 $. It is possible to connect two simple BOBs on it (one BOB ~ 12 $). Two BOB’s have 10 inputs and 24 outputs. The CNC-Software is running on a standard Win7 PC. The software license costs 68 $. Speed is no problem. The Ethernet connection allows up to 400 kHz operation. Note: Working via parallel port is much slower. Look at Google and search to “UCCNC”.

@Peter_Spiess ​ Be sure I will!

@Peter_Spiess “That’s not just my opinion, many friends agree with me” so it’s the opinion of you and your friends?

I was looking for a verifiable reason, that Ethernet is better.

But you seem to have a whole raft of things bundled into ‘Ethernet is better’ that have nothing to do with Ethernet being better.

Generally, faster is better, but once you reach the limit of the speed you actually need, any increase doesn’t improve anything.

I work on commercial CNC machines that have parallel-port connections and operate at 10x the speed of my home CNC - the 1.1m/bit/sec that a parallel port can deliver is not really a limitation, when you consider what the machine needs - an instruction to go to the next location.

I send my gcode via Bluetooth and WiFi, both significantly slower than Ethernet and yet they both work admirably.

I have two motion controllers with Ethernet and it never gets connected.

I don’t doubt that Ethernet is convenient, easy to locate, has enough speed to send text data 250 bytes at a time and is reliable, but so are serial, parallel, USB, WiFi and Bluetooth.

You haven’t made a compelling case for ‘better’

@Alex_Paverman I can make a ‘BOB’ (which is, I assume, a controller capable of decoding gcode and driving steppers?) for only a few dollars. There are boards like Smoothie, TinyG, and any number of GRBL-compatible boards you can buy for under a hundred bucks. If you have external drivers for your steppers, you could get away with as little as $30-50.

There are, it seems, two schools of thought on home CNC - those that use dedicated PCs to drive their machines and those that use dedicated, purpose-built pcbs and distributed control software .

My CNC laser and router are controlled with a repurposed Android phone with a big screen I bought for US$65. Their motion controller boards are a TinyG and a smoothieboard and the control interface is chilipeppr - a web/browser-based app.

The days of dedicated computer workstations running monolithic code are disappearing as specialised multi-core 64-bit processors have plummeted in price and boutique pcb manufacture has blossomed.

@Peter_Spiess ​​ I looked at UCCNC and that doesn’t bring me too far away from my ‘old’ conception. OK, I might have missed the price range of the solution, still 10x expensive than a simple parallel BOB. For the rest, it is still a computer directly attached to the machine as +Mike Thornbury says. I cannot compare that software with MACH because I simply don’t know it. Though, reading briefly the correspondent forum, I detected some possible ‘complaints’… But I dont judge what I don’t know by myself. Anyhow, what I have ‘learned’ about is that exists such way to work. It is better or not I can’t say. Only it is more expensive, maybe not so far away as I expected. For myself, I cant’t find a good reason to migrate to such solution.

@Mike_Thornbury ​ Not only I totally understand the concept, I am even using it! But only (for now) to my 3d printers. But you know why? Because of the software. It is not so simple to make a CNC software as complex and performant like MACH, working ona PI.
In fact, I am studying now this ‘branch’ with an OrangePI board for using with a (simple) CNC lathe I am building. Anyhow, as you said, this isn’t a “network” solution like Peter’s one