Advancements in automation

Members do not see these Ads. Sign Up.

The machine in this video is pretty much running through basic tests of my micromill code, using the lm628 based controllers from techno-isel running on linux. After 2 more segments of code are implemented, it'll be adapted also to produce orthoses, and ultimately, will have a 4th axis added to it to manufacture orthotics automaticly with 0 human intervention after inserting and clamping down material. I'll also be using it to make the jig and base plate assemblies, reducing expenses for jig manufacture by no less than $1000.00 per jig, so yes, it's been a really good day. :drinks

 

Joe

What are you showing here that is different from using existing CAM software?
Is you system significantly cheaper or better than traditional M and G code systems?


Cheers

Phil

 

The language I designed for machining is generic parts is in a human readable format designed like a machinist would understand better. Additionally, G-codes genericly don't support edge finding or proximity sensors and on machines that do, it's generally a more cryptic definition.

In addition to any other machine that uses the Acroloop series of controller, is now also compatable with techno-isel's hardware, which means my native milling software will be able to work with considerably more machines, instead of converting the milling data to g-codes. Milling an orthotic with sofware designed specificly to make orthotics is also generally more friendly for setup and resuming of orders. My next targetted controller will be the galil controller, of which, I currently have their 10-base T controller in house for that task. The 'boring operation would be a good indicator that communication layer is ready for use, as the updates are every 0.010" (1/4mm) and the machine is running smoothly even at that tight of a point to point resolution. On a more positive note, having the machine in house gives me the ability to do one of the things my customers have requested, to add order information to the orthotics themselves, rather than to inscribe or label them. Of key importance in this area is the proximity sensor, which is calibrated to determine the length of each tool to be used, +-0.001", which combined with the topographical data of the orthotic can be used to scribe the order number right into the parts without weakening the structure of the device.

 

Just another video of some of the features I build into my software, which will enable me to allow my customers, at minimum, to produce some of the components for my scanners. and to quickly setup a new jig without having to guess.at alignment.

 

Well Joe,

Let me be the first to warn you that you may be visited by a friendly podiatric cyborg from the future hell-bent on wiping out you and your creation in the knowledge that this is what started the RISE OF THE MACHINES!

You don't have a ped-egg do you?

:empathy:

 

Joe

Again my question is why? (Genuine question as I am very interested in improving things and not maintaining the status quo)
Processing systems like Heidenhein are excellent at 3d work and have excellent 'look ahead' properties to keep things smooth and effective. They can also 'flip' jobs over to minimise user input etc.
The interpretation of modeling out put to CAM to M code is all automated and pretty simple (once set up).

What is different in your approach?
To make me move to another system there would have to a financial benefit - time savings etc.

Cheers

Phil

 

One of my goals in this project is to be able to decentralize 3D scanner production, provide custom styles of scanners to the labs I work with, and then, to give them the option of producing most of their custom design in-house. While there are specific portions of the scanner that must be machined by someone with knowledge of flycutting, the majority of the scanner can be produced by the same machine operator that operates their machines for making orthotics. By supporting both Acroloop controllers and Techno-Isel controllers, it means all of the labs I work with can help to expidite delivery of scanners, or at minimum, reduce international shipping expenses by producing the majority of unit in the country, and allowing the lab to complete assembly within their own country. As for why G/M-codes are so archaic, it's very limiting performance wise as well as being cryptic in terms of programming.

Well, there's that. I haven't showed my patented milling path yet. As a rule, it's twice as fast as milling laterally, and about 85% faster than longitudinal milling, on the same machine.

Here's the abstract

I was just showing some of what I've been up to. As of now, I have 2 systems to install out of the country, and 1 or possibly 2 systems in the country to install.

 

Well, the machine had a flaky encoder on it that I finally fixed until it broke, but you'll see more soon. Renco originally quoted the encoder at $200.00 The next stage, once the machine is up is to setup the automatic tool changer, proximity sensor, and depth gauge calibration holes. After that, interfacing to my milling files directly will be the easiest part of it all. The fun part to this machine is the jog speed, which is in excess of 700IPM and contour milling speed of 300 IPM, which means milling time per orthotic of under 1 minute per surface. As soon as I have the machine back up I'll put up a youtube video of its performance.

omg, I see those things at WalMart, which in this area is one of the only social events.

 

Just an FYI for those of you running labs with Techno-ISEL orTechno-CNC routers. It would be wise to inspect the thrust bearings for each axis and verify at minimum that they are angular contact bearings and not cheap bicycle bearings. My own router, as shown by the previous videos started developing backlash and after discovering the kind of bearing that it used , I ordered replacements from Vxb.

The original part,6200rs should be replaced with 7200B, for example, and will result in tighter tolerances.

What gets me about the Techno line of CNC routers is how they can put $2.00 bearings into a $20K machine. I wouldn't be able to look at myself in the mirror if I did that. For $6.00 more per axis, the bearings will both last longer and provide greater accuracy, and it seems a shame to not spend the extra $6.00 per axis.

 

There is a new video on my playlist that shows my patented method of producing foot orthoses. It has the advantage of continuous feed machining, instead of accelleration to a constant speed, decelleration, stopping to change lanes, etc. In addition to actually making orthoses, I am also designing my software to allow a regular end user of the product to make repair components for itself, effectively making it a self-repairing unit, for much of itself. I still have to align my machine, but afterwards, will make available photos of surface quality. This current experiment involves reducing the ball nose endmill size to 1/4", and seems to have been successful. This means an improvement in shape for all devices, and in particular allows for a much steeper climb into devices that need it, such as the UCBL design of orthotic. The caviet to this experiment is that the spacing between machining passes needs to be reduced for a smoother finish, but it does in fact have the advantage of saving 10% on material costs, allowing for 9 to 10 feet on a 12x24" sheet of plastic, instead o 8 to 9, using a 1/2" endmill or larger.


Note: In this example, I'm using a 2 1/4 HP router motor, which is obviously overloaded and currently running at 250 IPM (Inches per minute). In the video, you will also note that the spindle slows down in steeper cuts, so the next experiment will be at 200 IPM.

 

Another video part 1 and part 2 with better lighting, same channel.

 

New Videos, using a 1/2" ball nose endmill and 15 degree conical endmill

Part 1 - Top Surface contour machining
Part 2 - Bottom Surface contour machining
Part 3 - Part cutout