Should I build a big CNC router?

I know I would like to build a Computer Numeric Control  (CNC) router big enough to drop a sheet of ply in, but is it a good idea? It will take up a lot of my workshop space, and chew up a lot of time and money. It’ll allow me to makes some awesome large objects. Anyone want a kayak? It would be big enough to make small boat ribs etc. Storage boxes, and display panels, doll houses and 1/2 of the things in ikea. All doable. But am I going to do them? A lot of the CNC routers on the internet seem to get used to make more CNC routers. Thats fine by me, although a boat sounds more fun.

I’ve read about people who have purchased shopbot routers with the belief that they’ll form some kind of business around them, and some have successfully done so. It seems a bold choice to me, but as James Graham said.

He either fears his fate too much,
Or his deserts are small,
That dares not put it to the touch
To gain or lose it all.

Maybe now is the time to build a big CNC router.

PS. I’ve found a few places around that I can get stuff cut at. I’m going to test the waters out by doing a few projects on someone else’s CNC router first.

Caking with terror

I decided to make an upside down banana toffee cake for my wife for her birthday, because she’s awesom. I’m not the best baker, so I normally stick to easy recipes. But in a fit of bravado I decided to stretch my self.

At 4pm on a Tuesday, while making roast lamb and chasing two kids.

I’m not sure why I decided that was the appropriate time to branch out, and in retrospect it doesn’t seem like the best choice. But the cake was delicious.

I didn’t have any problems with the cake side of the recipe, it was the toffee that freaked me out. That stuff is just dangerous. I’d feel safer making napalm. I got the roast lamb on, started the syrup boiling, made the cake mix with one eye on the syrup.
I now know that if you hear the boiling become quieter, it’s a sign that the toffee is about to start browning. Unfortunately, with the first batch, I wasn’t aware of that. I turned around to comfort an upset 4 year old who’d bumped his head, and turned back to find white smoke
coming off the toffee, which was now black.

I got it off the heat and left it to cool (NB: molten sugar is damn hot, don’t put it near water. Even though it looks still, it’s still well over 100C. Any water will flash to steam. It may blast volcanically hot toffee everywhere in the process.) I cranked the exhaust fan up and grabbed a new saucepan.
This cake was going to get toffee. My son had decided that a kitchen that had stuff with smoke poring out wasn’t where he wanted to be, so I got 15 minutes of peace to try again.

The second batch went well. I followed the recipe. Stirred while the sugar was dissolving and kept it below a boil, didn’t stir while it was boiling. And i didn’t do anything else while I made it.

I did give it a little stir as the sugar began to brown as the colour was uneven. I didn’t realise how much the toffee would darken up after I stopped heating it, so I almost killed the second batch too.
After pouring the molten toffee into the bottom of the lined cake tin, you add a layer of banana slices. You can see them start to boil as they go into the toffee. If you weren’t motivated to keep your fingers clear before, you should be by now.

It all worked out in the end. Dinner was only a little late. The roast lamb was very nice, and the cake thick and strongly flavoured by the banana and the tofee.

Camping roas lamb


  • 1 leg of lamb that will fit in your camp oven
  • 2 cans of crushed tomatoes
  • 1 can of water
  • 1/2 bag red lentils
  • 4 large onions
  • 2 cloves of garlic
  • one tub of vegetable stock
  • olive oil
  • salt and pepper
  • camp oven or large cast iron pot


  1. peel the onions and cut in half sideways (so the onion rings are visible) and line the bottom of the camp oven with them. They should pretty much cover the bottom.
  2. place the red lentils on the onions.
  3. place the garlic on the lentils.
  4. place the lamb on the lentils and garlic.
  5. give the lamb a dash of olive oil and rub it in. season with salt and pepper.
  6. pour the tomatoes either side of the lamb
  7. mix the stock into the water and pour half either side of the lamb

Cooking (in the kitchen)

  1. pre heat an oven to full whack (as Jamie says)
  2. stick the pot in with the lid on
  3. after about 15 min turn the oven down to 180 (160 fan forced) and cook for 1.5 hours
  4. take the lid off and inspect, I left the lid off to brown the top of the lamb at this point. Just keep an eye on the tomato to make sure it’s not burning.

if you would like to do this as a slow roast, just leave the lid on and increase the cooking time to about 3.5 hours.

Cooking (in the fire)

Place the camp oven in a bed of hot coals and scoop a shovel full of coals onto the top of the oven. Check it every half hour to hour to make sure its cooking ok.
We left it to cook for 3.5 hours. Instead of carving the roast, we just pulled it apart with forks.

iPhone controlled paper planes!

I’ve long been a fan of paper planes. Some of my fondest childhood memories are of are of constructing and launching paper planes around the school yard.
For a while now I’ve been eyeing off the powerup paper plane power pod on thinkgeek. Debating if I should get one or if I should just stick with the paper in its
true form. Then I saw this on kickstarter. Remote control paper planes! Convert any paper plane into a remote control plane and steer it with your iphone. Its a two channel control system. If you want to climb, then throttle up. It turns through rudder (Yaw) control, so there is no roll control.

This is the minimum you need to control a plane. With the control unit and battery up the front and a propeller and motor at the rear, the body looks like a
thin black stick. You lay the system down the body of your plan, throttle up, and fly away.

For $45 deliverd to Australia, I’ll give it a go.

I also found an interactive bokken training sword on kickstarter. Its possible I’m addicted to kickstarter, but I’m resisting the sword. For now.

The Mini Mill is live

I picked up a 7×7 mill from Zen toolworks a while ago. I did the basic assembly and powered it up with an old 24v power supply. It moved, but locked up on the far X axis, and the far Y axis, and stalled with any high speed movement. At that point it all got boxed up for a house move, and stayed that way for a year. I’ve recently dragged it back out and thrown some more time at getting it setup.

I lined up the bottom frame using a machinist square. It was close enough that I could just shim out one corner with a piece of paper. The stepper motors  pull more power the faster they run. The old 2.5 amp 24V psu wasn’t providing  enough power as the acceleration kicked in. I’ve swapped it out for new 14.6 amp 24v psu. I still found some issues with the alignment of the anti-backlash nut on the Y axis. The problem with this is that you can’t easily access the anti-backlash nut. After a good look around the machine, and I decided to remove the 16 screws securing the back plate from the gantry. I could then access the mounting screws for the anti-backlash nut by loosening the  mounting screws and moving it to the far side of the Y axis. I could then tighten the mounting screws with the anti-backlash nut in the correct alignment.

Also, I got a video cam for my birthday, so here’s a time lapse of the job progress.

Northern Makers meet up

Northern makers had a meet up in Westgarth on Saturday night. It was a bit late notice, but we still had five people there. There was a Mendel Max being assembled, some arduino controlled led strips, analogue synth boards being soldered, and I was trying to work the kinks out of my mini CNC mill.

They are a nice bunch of people, and it was very good to see what other people where working on. They’re proposing to book it in as a regular thing, one Saturday a month. I suspect it’ll be a regular outing for me.

Quadcopter update

I’ve been a bit slack on my updates, so here’s a couple in a row.
I got some time to fly the quad copter on the Melbourne Cup weekend, while I was at Apollo Bay. There wasn’t much of a window between the weather to fly the quad. I’d been playing with the settings on the controller and I had calibrated the ESCs. The result was a cascading build-up of oscillation. It was messy. At one point I tried to keep it flying hoping it would calm down. It was not a good plan, the quad completely inverted and smashed into the grass. I managed to land it two out of three times, but by then there were a few people coming over to look. I wasn’t comfortable flying around without adequate control of the quad, so back into the box it went.

I’ve been doing some research, and it looks like I’ve set the ‘I’ values to high in the PID loop. I’ll set them lower for the next run, which will be tethered to the ground to reduce the risk of damage. I’m going through too many propellers.

Quadcopter Update

Well, I’ve learnt the difference between ‘Acro’ mode and ‘Self Level’, and it only cost me one propeller.

This afternoon after work, the kids and I took the quadcopter to the local oval. It started out as a lot of fun, but I hadn’t realised that you can change the mode by stick input on the controller. Somehow I managed to knock it from Self Level mode (I need all the help I can get at the moment – I’m up to almost 6 minutes of flight time), into ‘Acro’ mode. I’m assuming ‘Acro’ is an abbreviation for Acrobatic, because it got very exciting after that. If you are expecting the quad to self level when you center the sticks, it takes a moment to realise that it isn’t happening. You don’t have a lot of moments available.

Here is how the construction has progressed. I’ve completed a bunch, and learnt that I have more to go. It’s flyable now (or it was before this afternoon’s adventure), but there is lots of room for improvement.



  • Install video cam
  • Install video TX

i’m not 100% sure what has gone wrong here. I wired the camera and the TX together using the supplied wiring harness and powered it up. The video TX blew and managed to ignite the heatshrink. I suspect its taken out the camera too. My best guesses are faulty wiring or faulty board. Without better diagnostic equipment, there isn’t much I can do about it. I’ll have to order replacements when funds allow.

To do:

  • Replace the rear props with some red ones. Its hard to maintain orientation with the yellow front and white rear. There isnt enough contrast. Also, I broke the white one. 🙂
  • Re-mount the KK2 and re-level it. It tore off two of four pieces of mounting tape in the impact.
  • ESC Calibration (electronic speed controllers)
  • Instal SimonK firmware on ESCs, I’m leaving this until I have a rainy day
  • build (anti)vibration isolation platform

Password storage in .net

Password storage is a pet peeve of mine. I’ve worked with systems that use:

  • bidirectional encryption for the passwords (you can convert the password into gibberish, and then back to readable text),
  • MD5 hashing (you can convert the string to gibberish, but not back. You test the password by hashing the password that the user is trying to log in with and comparing it to the hash in the database.), and
  • clear text storage. (This is the worst plan ever).

They each have their own security flaws. Bidirectional encryption can be converted to clear text, which means that if someone steals the user list, they can discover all the passwords. Likewise, you can build a dictionary of known passwords and MD5 hashes and use that to look up the password that made a hash. And storing a password in clear text is pretty much the same as just giving out your user’s passwords.

I’m currently working on a software application that uses the BCrypt hashing algorithm to store passwords. BCrypt is available for .net and other languages, and is very easy to use and implement. The BCrypt library produces a salted hash, and stores the salt and the hash together. This systems means that even if you had a dictionary of hashes, you would need to rehash your entire dictionary for each salt. As each salt is unique, this is going to take a long time. The algorithm is intentionally slow, but configurable. By default it runs about ten iterations, but can be adjusted higher. The trade off is between security and hardware. In theory, as the hardware gets better you can increase the number of iterations, and keep it difficult to compute. The estimated execution time for the default configuration is 0.3 seconds per hash.

So, to summarise, BCrypt is:

  • a one way hash (you cant reverse the password out);
  • uniquely salted, making pre computed dictionary attacks much less useful;
  • adaptive, so it can handle future changes in available compute power; and
  • easy to implement.

So people, hash your passwords!

More information at