28 August 2015
27 August 2015
26 August 2015
For too long I’ve thought and talked about doing some posts of things I’ve made. After all my blog IS Continue reading
24 August 2015
23 August 2015
|60 of these go into a full sphere|
20 August 2015
These were very handy at the weekend just gone. It's a map of the quarry we play Vale LRP in, we added named locations so that everybody looking at the map was talking about the same location. It was actually really handy and I'll be making some more of them in the near future.
18 August 2015
14 August 2015
A non laserable in preparation for an upcoming Vale Event. This Scale mail gauntlet is made up of 200 individual metal scales all weaved together with 400 smaller split rings. It's a pretty cool piece especially considering I had the bits floating around in the garage for several years now. Scale is definitely harder to make than chain but looks prettier.
13 August 2015
After the cabinets came the keyrings, neither of these was the winning design but they're both pretty close. As I didn't take a photo I guess you'll have to see Simon at an event to find out what the final design looked like.
12 August 2015
All Sharks machines come with a honeycomb and a knifebed; the knifebed is usually stored under the honeycomb because lifting the honeycomb up prevents smoke from being trapped under it. While I was cutting some more Clarkcade gaming cabinets I realised that I should be using the knifebed instead of the honeycomb. It's perfectly suited for this kind of job where the panels are much bigger, All the shrapnel falls through the gaps allowing me to put the next sheet on without even cleaning up and the cuts were all much cleaner due to the reduced number of reflections.
|6 new Clarkcade Cabinets ready for postage|
11 August 2015
09 August 2015
Monitoring the rotational speed of the wind turbine blades can help with power performance measurements.
This post shows various techniques to measure the rotational speed of a 'wild AC' output small wind turbine. In this system, the wild AC is rectified into DC and either delivered to a battery or to a grid-connected inverter.
The wild AC output of the turbine can be used to give us the rotational speed.
From November 2014 until May 2015, I have been one of four "Makers in Residence" at Derby Silk Mill, in the UK.
This has been a slight change of activity for me, as I usually focus on renewable energy systems, but has been a great opportunity to encourage Science, Technology, Engineering, Art and Mathematics (STEAM).
I hope to encourage the younger folk (and all ages!) to get involved with future engineering and technology issues.
I kept a separate blog of my work at the Silk Mill, but here are some details of the work I have done here.
06 August 2015
Manchester Makefest is going to be running this weekend and one of the collaboration projects they're attempting to run is Amazeballs. Several people build modules that contain marble runs for 27mm pinballs, the modules are brought together at the faire and the machine all works (in theory).
Marble machines being kinda my thing and a desire to help people with interesting ideas Just Add Sharks ended up signing up to make 2 modules for the machine. Dominic has made one and here is my module.
My first attempt to make something fancier failed completely, the balls are pretty big and heavy and sheer momentum was smashing the machine apart during testing so I went back to the drawing board and came up with this passive machine. The 68mm drop between start and finish isn't a lot to work with but it is possible to make something interesting that can run that length.
03 August 2015
02 August 2015
Recently I’ve been focussed on finishing off my radio controlled model SpeedTwin ST-2. This guy is not the biggest model I’ve built in terms of wingspan, but it wins in terms of chunkiness and complexity. I started it in 2012, but it stalled at some point because everything was blocked by scary “one shot or it’s ruined” style tasks. I recently dug it out and decided to get these things over with so I could get it back on track.
The fuselage was stalled as the servos and radio receiver needed fitting before any progress could be made, as they would soon be difficult or impossible to access. After fitting the servos I coated a sheet of 1/16th balsa with epoxy and stuck it to another block coated with tape, shiny side out. Once the epoxy was dry the taped sheet was removed leaving a perfectly smooth surface which the receiver could be stuck to with double sided foam tape. I glued this mounting to the side of the fuselage with a few balsa rails so it could be easily removed if necessary but was still securely fitted. This was mounted behind the canopy area in a spot that was still accessible while being as far as possible from interference from the battery and speed controllers. This location also provided a convenient spot to mount the antenna vertically.
With the servos and receiver installed the decking on the rear of the fuselage could be added. This is designed to form a skin over the formers and spine of the fuselage and give it the correct final shape without having to use excessive material or sand material away. To fit this I first figured out the best way to have the grain match the curvature of the surface and cut some 1/16th balsa to shape. After a few test fits this was glued where it meets the fuselage sides and left to dry. It was then curved around the formers and glued a bit at a time to avoid overstressing the material. Despite this a crack formed at the very rear. I was able to fix this by glueing and pinning it in place and sanding it smooth later. Once this was glued on I cut the sheeting down the centreline and repeated the process for the second half.
With the rear of the fuselage essentially finished I moved on to the front end. The nose was built up from blocks of 1/2 and 1/4 inch balsa which were glued on and then sanded to shape. This process requires a lot of sanding allows for nice curves to be shaped easily. The canopy was built in a similar way but from a block of blue foam glued to a balsa base. The foam was shaped to the correct profile with my hot wire cutter and then sanded the rest of the way. A 1/64th ply windscreen section finishes it off. The canopy base has some rails glued to the bottom which locate into the opening on the fuselage. The canopy is then held in place with a dowel at the back and will eventually have a catch installed at the front.
At this point the fuselage was basically finished except for the covering. I decided to use a lightweight fibreglass technique for this with 25gsm glass cloth and water based polyurethane instead of epoxy resin. An internet search will provide plenty of good instructions on this process so I won’t expand upon it here.
With the fuselage out of the way there still remained a lot of work to be done on the wing.
First the leading edge of the wing (formed from a strip of pre-shaped balsa) was glued on to the front of the wing and sanded to match the wing’s taper. With this in place I was able to locate the wing in the fuselage and measure from the tips to the rear of the model. Once these measurements and the measurement from each wingtip to the fuselage matched I was convinced the wing was aligned correctly. I drilled a hole for a locating dowel into the front of the wing through a hole in one of the formers and another hole through the wing so it can be held in place by a nylon bolt. The dowel and bolt will provide a secure and repeatable fit for the wing (but hopefully the nylon bolt will break instead of the wing in the event of a crash).
The next part of the wing that needed attention was the engine nacelles. The original has huge engines hidden inside bulbous nacelles which needed to be recreated (even though my motors are relatively tiny). I chose to use the designer’s recommended method for this which is to plank the nacelles by glueing strips of wood over formers, with some foam parts where the curvature made this impractical. I decided to build these in two parts with the upper part permanently attached and the lower part removable in case access to the landing gear mounting points was needed later.
During this process I ended up making a second hot wire cutter for making parts that needed square edges (at least before sanding) and for parts that needed a consistent thickness. This consisted of a frame made from an old speaker cabinet with a hole drilled in the centre. A screw inside the hole mounts the bottom of the hot wire which runs to a similar hole at the top. The top of the wire is mounted to a spring (for tension) attached to a block which can be moved about and clamped in place. Spacers under the block that holds the top of the wire allow adjustment of the sprint tension. A thin strip of wood attached to the table with a smooth shank screw makes a fence which can be clamped in place to cut parts to a consistent thickness.
The planking was slow and tedious but performed in short bursts – adding a few strips and leaving the glue to cure while doing something else. Unfortunately I wanted to build the upper nacelles first, so the removable section could be built to fit them. I learnt partway through that using softer wood was better for this process and the lower nacelles came out nice and smooth with just sanding. The upper nacelles made from harder balsa required quite a bit of filler and some reinforcement from beneath to get them nice and smooth.
My motors will fit on the front of the part of the nacelles shown above, with an extra bit of foam that will be sanded to shape to hide the motor and provide a nicely shaped front section to the nacelle.
Once the nacelles were built I installed the landing gear struts using some P shaped clips made from brass. These were made by folding brass strip around the landing gear wire using a vice, drilling a hole in the correct place then trimming them to size. This allowed the struts to be bolted to plates attached to the underside of the wing. A hole had to be cut into the lower nacelles in a suitable place for the strut to fit through, with some clearance to avoid damage if/when the wire flexes. Cutting these holes was a bit nerve wracking after spending so much time on the nacelles, but after measuring several times I was able to hit the correct location first time (with some extension of the holes to fine tune the fit).
Finally the wing tips were cut from from blocks of balsa, roughly shaped and then attached and sanded to their final shape.
Once the wing was complete I did several passes to check for dents and other issues (which were fixed up with filler) before glassing it in the same way as the fuselage. The nacelles were glassed first followed by the rest of the wing, with cut outs around the upper nacelle area to avoid the extra curvature causing problems.
Glassing the fuselage added about 30g to the fuselage (which originally weighed about 235g) and 50g to the wing, which originally weighed about 495g. This is not a problem and well worth it for a sturdy finish. As the numbers show, most of the weight and complexity of this model is in the wing since it’s a twin engine design.
This brings the project pretty much up to date, and all that remains is the last 10% which will probably take 90% of the time!
01 August 2015
31 July 2015
When doing some research on the web about my Philips X.40 electronics teaching kit I found out about the contemporary Philips EE2007 kit which had a battery operated CRT for the teenage nerd to play with. Just how cool is that? Even as an adult I immediately wanted something similar. The schematics of the EE2007 CRT module are easily available and I thought about building a clone using modern components but, to be honest, I am a bit of a scardy -cat when it comes to high voltages and I didn’t fancy getting to grips with the HT supply. Like so many others the project went on the back-burner.
Recently though I was thinking about upgrading my boat-anchor oscilloscope (it’s a 1980s(?) high-spec Hitachi hybrid analogue/sampling scope which is a joy to use but massive and very limited by today’s digital standards) and I thought I might be able to get a second hand modernish digital scope on eBay without spending too much. I didn’t find the ideal match there, but I did see a Thandar SC110A miniature analogue scope. It was love at first sight – a battery powered scope in a small, smart enclosure that matches my function generator which is always on my electronics bench. Some people find kittens cute but, for me, this dolls-house scope with its tiny 4cm diagonal screen just makes me feel all squeeeeeeeeeee. In XY mode I could use it like the EE2007 CRT but with added convenience of calibrated amplifiers and a timebase (no intensity input though).
The scope was described as working but I was always sceptical because eBayers are often economical with the truth and anyway not many people would know how to test it properly. Without being cheap it was at a price that I was prepared to spend for a bit of fun so I bought it expecting I would have to do some remedial work. When I got it there was good news and bad news. Good news was that you could get a “bright line” trace with no signal and most of the functions seemed to do vaguely the right thing. Bad news was that the display had awful distortion and freaky interactions between the X and Y axis. Also some of the input ranges on the Y axis didn’t seem to work. Time for fault-finding and recalibration using the handily supplied service manual.
To do the calibration you turn the scope upside down and take the bottom half of the case off which exposes the main circuit board and calibration controls. Once inside you see that the circuit board is actually branded Sinclair so you will understand if I was suspicious that it could ever be made to work properly at all. In fact Thandar and Sinclair seemed to have cooperated on several test instruments. Sinclair seems to have picked up the case mouldings from Thandar and Thandar sold several Sinclair instruments under their own brand (brave decision). This design seems to have originated in Sinclair and the CRT, as we will see, was custom made for Sinclair for their microvision pocket TVs.
When I powered the scope up to calibrate it I realized that it had a fault that I had never seen before in any device. It worked, to an extent, the right way up but when you turned it upside down the trace disappeared and the whole thing seemed dead. I spent a long time turning the machine one way and then the other trying to find out where the thing stopped working and where the fault might me. Eventually I heard a little tinkle that seemed to come from inside the CRT which alerted me to the real cause. Sometimes electronic debugging involves all your senses. Removing the screening foils and looking inside the CRT I found that the glass supports holding one of the Y axis deflection plates and place had broken it the electrode was just flopping around – coming in to contact with other electrodes when the scope was turned upside down. Humm – eBay seller was dodgy on that one IMHO. Still, at least we have a plausible explanation for many of the problems.
Now, where do you get a custom made Sinclair CRT from in 2015? A few specialist shops had them listed at silly prices and were out of stock anyway. Then I had an inspiration – my searches had told me that the Sinclair Microvision TVs used the same tube so perhaps I could get one of those and cannibalize it. Back to eBay and, yes, there was a suitable Micovision for sale at not too high a price.
While waiting for the Microvision to arrive I did the best to fix the other problems working with the dodgy tube. To do this I had to strip the case off completely so I could run it the right way up and still access the controls. As a matter of routine I decided to replace all the polar capacitors (tants and electrolytics) in the design. It has a switching power supply which is nice and compact and gives it a very wide tolerance of supply voltages. In my experience switching power supplies were pretty rare at the time so this might have been a brave design. Normally in test equipment you see a lot of effort going in to conditioning the power supplies with great big capacitors to smooth them. In a sign of classic Sinclair cost-cutting the biggest supply capacitor in this scope is 47uF. It seems to work well enough though.
Next I turned my attention to the dead ranges on the Y Axis. . I spent a long time investigating possible faults with the switch or problems with the Y input buffer. In fact I should have checked the obvious. There is a “DC Offset” trim which needs to be set right to get the input buffer working correctly on all ranges. RTFM applies here.
After a few days my Microvision arrives in the post. I have to say it does look like a proper gadget and must have been quite a cool thing when it was new. As far as I can tell, with no suitable source for a modulated analogue TV signal available, it worked fine. Inside it’s in good condition and there is a surprising amount of space in the case. I feel a bit bad about whipping its tube out and swapping it for the broken one from the scope but I reckon it’s better to have one useful device that works than a TV that is no longer usable in the digital era.
With the new tube in and the polar capacitors replaced the scope is working a lot better. Almost everything seems fine except that the Y signal is still very distorted. My fear at this stage is that the shorted electrode in the old tube has damaged some components in the Y amplifier and it’s going to need a lot of debugging and component sourcing to sort it out. In a moment of inspiration I wondered if looking at the lissajous figure from feeding the same signal to X and Y might be useful to understand what was going wrong. Instead of the expected straight line I saw a very definite knee – like some part of the amp was saturating. It turns out that setting up the calibration for the Y amplifier involves two interacting gain controls. The previous person to set this up had set one way too high and the other way too low forcing the amp in to saturation and clipping. Once both had been reset to near their midpoints the amp worked just fine. To be fair the service manual doesn’t explain how to set these controls very well and it requires a bit of electronics experience to know what you are supposed to do.
With the Y amplifier recalibrated everything now works very nicely. Technically it’s certainly not a great scope but its simplicity is lovely. The tiny size is very practical. I like the immediacy of analogue scopes. You interact with them much more by intuition rather than digital scopes where you have to think. For answering the eternal electronic engineering question, “do I have a signal here?”, on my audio circuits it will be very handy. As I nice final touch I made a lead so I can use a USB battery booster that I got as a freebe from a trade-show as a power supply. It should run for about 10 hours on a charge. It’s just a really nice thing! I’ll maybe think about whether I can add an external control for the intensity and perhaps do a DIY TV on it.
30 July 2015
29 July 2015
I hadn't intended to do another coffee table for some time but we had Lego spread all over the coffee table and I was getting bored of rummaging around the bottom of deep storage boxes so I decided we needed some real Lego storage and lots of shallow trays to make it easier to find the parts. There was also a convenient gap under my Escher coffee table so I figured I would base it around another Lack coffee table.
We started to fill the trays up with Lego and we realised just how much Eli has already, once I get my Lego into the mix I think we might need a second storage system. At least I have a happy customer for the time being.
28 July 2015
27 July 2015
The new useless machine design has a PCB which allows it to be assembled easily, it also allows me to sell machines preassembled. The MAD museum are selling the 'No Soldering' version of the machine which means I end up soldering the boards myself. I remade this jig to hold all the boards in place, this post is mostly about the continual improvement though. The last iteration of this jig had round holes which let the PCB's twist in place, this one has square holes that key into the switch and hold them tight. The joy of having everything laser cut is that you can refine these things over and over again with minimal cost and effort until it is exactly right for the job.
26 July 2015
19 July 2015
"Have you noticed that the more frequently a particular open source community tells you to RTFM, the..."
“Have you noticed that the more frequently a particular open source community tells you to RTFM, the worse the FM is likely to be? I’ve been contemplating this for years, and have concluded that this is because patience and empathy are the basis of good documentation, much as they are the basis for being a decent person.”
- Rich Bowen (via roomthily)
16 July 2015
With the style decided upon but me headed away on holiday, I made a final tweak of the artwork and sent it up to Dominic at Just Add Sharks. He did the actual cutting for me and delivered it to Totally Brewed in time for their event. It looks great on their bar.
15 July 2015
I've been designing a breakout board for the ACS758 hall-effect current sensor. The first few prototype boards came in and I decided to give them a quick test. The results were not quite as I expected so I thought I'd put write about them here, for others that might be using this IC.
These current sensors come in a number of ranges (50A,100A, 150A and 200A, all with uni and bi directional variants). They use the hall-effect, which measures the change in the magnetic field to measure current. This means you do not need to install a shunt resistor, with its associated power loss and voltage drop.
14 July 2015
My new aesthetic is KNIT GOTH
07 July 2015
06 July 2015
Or why a granny square crochet afghan is ubiquitous.
I can’t crochet a granny square, but I do keep on meaning to do something like this.
05 July 2015
Ed Saperia, inventor of Heredox and a million other cool things asked me to make a large version of the Heredox tile game for garden parties. It's made from 12mm exterior ply that was cut at B&Q on their big saw. The symbols however were laser cut from 3mm poplar and affixed to the top of each tile, with laser cut markings on each tile to show where the symbols should go. The whole game was then strapped up into a carry case which has the rules engraved on it. As a project it took way longer than expected but it's a really cool item I got to make so that's pretty rewarding.
01 July 2015
I just cut a batch of gaming tokens for Tree, these are apparently for the game "Guild Ball"
29 June 2015
24 June 2015
19 June 2015
The Minecraft coffee table is now up and visible. There are lots of pictures and a detailed description of the build process available over on the Kitronik Blog. This project idea lingered for a good 12 months and having the support and backing of Kitronik really made it possible to put it together in super fast time. Go check it out now
14 June 2015
09 June 2015
08 June 2015
06 June 2015
The bins all stack on top of each other and a small overlap locks them into place so they won't fall off. The racking system is also versatile because it can be wall mounted or free standing, it can accommodate any sized bins in any position.
03 June 2015
A few weeks ago I headed down to rural France, near Toulouse to take part in a week-long Development Camp working on wind turbine measurement for the group Wind Empowerment.
About 10 people came from all parts of Europe, along with people joining via Skype from all parts of the world.
We discussed ideas and specifications and prototyped equipment for monitoring wind resource and remote monitoring and managing wind turbines.
This is a simple, low-cost amplifier kit based upon the LM386 amplifier IC.
It has a maximum output power of 1W, which is good enough for amplifying mp3 players, phones and also the beeps and sounds from microcontroller projects.
Please note: This kit has one surface mount component, the LM386, which is good for those learning to solder surface mount components, but not best for people new to soldering. The rest of the parts are through-hole.
It is available for just £6 (including delivery within the UK).
01 June 2015
27 May 2015
25 May 2015
23 May 2015
22 May 2015
21 May 2015
20 May 2015
19 May 2015
17 May 2015
16 May 2015
I wanted to make a little Baymax to go into Eli's birthday party invites. This is cut from 0.8mm white polypropylene. The details were coloured in with a black fine liner pen. (svg here)
15 May 2015
12 May 2015
11 May 2015
04 May 2015
I was at a loose end during today’s bank holiday so I decided to do a mini-project (with bonus recycling features) and make a set of control horns for my radio controlled SpeedTwin ST-2.
After deciding on a sensible size for the control horns I drew them up in DraftSight. These ended up being 25mm tall and 12.5mm wide at the base. I designed them on a sprue so that when I come to etch the remaining copper off they don’t get lost in the etching tank. The horns are shaped so that the holes for the linkage can lie on the hinge line with plenty of material around them for strength.
The DraftSight file for the parts can be found on GitHub in my rc-parts repository, which also contains drawings of various bits of radio control hardware that I’ve designed parts around. Please feel free to use any of it or to contribute drawings of parts.
The PCB material I used comes from a nice FR4 board onto which was otherwise unusable due to a design mistake. The board was intended to be used as a 2 player version of Charlie’s ‘Minigun’ miniature SuperGun project. Unfortunately after etching I noticed I had messed up the pinout of the JAMMA connector when transcribing the design from Eagle to KiCad. Rather than throwing it away this project is allowing me to reuse the material for something useful. My 2 player ‘MiniGun’ will eventually get finished and written up too.
As a quick aside, the SpeedTwin has made some progress since I last posted about it. The fuselage is nearly finished, except for sanding the canopy to shape and some more work on shaping the nose cone. The wing is also coming along, with the top half of the engine nacelles planked and mostly sanded to shape. The remaining work on the wing is to install the tips and build the bottom half of the nacelles, which will be removable for access to the landing gear.
Once the control horn design was ready I used CamBam to convert it to G-Code. Unfortunately some manual editing was required on the output to get Grbl to accept it happily. The main problem was a G17 code, intended to signal that arcs should occur on the XY plane, which caused Grbl to error after any subsequent G3 (arc) code. This setting was default anyway so the line was removed with no adverse effects. I also tend to remove comments from any code that is passed to Grbl – the parser can choke on lines over 50 characters so comments at the end of lines are best removed. If I find an open source CAM program that will provide Grbl compatible G-Code out of the box I will probably switch to it, I just need to put in the time to find one.
This is the first time using my eShapeOko since I rebuilt the controller so I had to spend some time setting up again. Once I’d calculated the appropriate steps/mm settings for each axis everything went fairly smoothly and after a few ‘air cut’ test runs I cut the parts. Since this part is all made in one cut I simplified things slightly by removing all of the Z axis movement from the program. I manually plunged the bit into the work from GrblController and then set the program going.
I decided not to drill the holes on the CNC to save setup time and because I don’t have a suitable drill bit that fits my eShapeOko’s rotary tool. They will be quick and easy to drill accurately on the drill press at NottingHack at a later date.
All that remains is to etch the remaining copper off (and sadly lose the current futuristic look), drill the holes and cut off the sprue.
02 May 2015
More Leather work for Marcus. I was trying to align the etch to the middle of the leather that he sent me but it didn't really work. Clearly I should have used the laser to cut out the leather as well.
01 May 2015
30 April 2015
- Add some character to your box, with a real time clock chip make it come out slowly after midnight or once every hour.
- Add some neopixels to make the inside of the box glow.
- Add a second motor to make it wave the flag of surrender or run away when it gets annoyed.