Please send shark

I have decided my next project needs a laser attached. Having bought a small laser pointer to butcher I then decided I really ought to just buy some laser modules as they can be had cheaply on eBay.

I can foresee these getting used a fair bit because even when they're just 5mW cheapies, lasers are still cool.

You wait for a bus

Then three come along at once.

I was showing my partner progress with the Time Machine at the weekend and it suddenly died.

As expected when I looked tonight this was down to crappy power distribution, something had come loose.

It's been obvious for a while that I need to sort this out so I'd already ordered some busbars to do it. For the wire gauges and amount of current involved they're OTT but they are nice and tidy.

I also suspect the tiny switched mode DC to DC converter I bought to generate 5V has died. Most of this runs off 12V and the Arduino makes 5V for the odd bit like the remote control but it's not enough current to power all those servos.

Nixie not

I have for some time intended to come up with some kind of instrumentation that you need to peer into on the time machine.

The idea is that it's something that will show you a readout of consequence without it being as modern as an LCD screen. To force the whole 'peering' thing I decided it would have to be seen through a security door viewer. Besides I happened to have one kicking around.

In a perfect world what you peered at should be something made from Nixie tubes but realistically that just isn't going to happen.

The next idea was an 8x8 LED dot matrix as they are cheap, very versatile and I have one kicking around, along with the serially driven driver chip to make it easy to use. I've been intending to go with this for a while but struggled with how to make it look good.
All the LED matrices I've seen have been pretty chunky things, designed to act as a big bright visual display. Smaller things tend to be done with LCDs, for understandable reasons.

So I've struggled with how to make this work physically. I was going with hollowing out a knob from a newel post but it was simply huge and started to go awry as soon as I started doing it.
Last night it occurred to me that going more old school and using a seven segment LED display would give me a much more compact option and keep the 'primitive' feel at the expense of being a less versatile display.

Some very quick fettling up of a plastic tube originally holding vitamin pills has turned out a pretty good looking viewer in my opinion.
Once you actually peer in it's pretty obvious it's not a piece of steampunk tech but until then it's a tube with a glowing display in and pretty much exactly what I wanted.

I may experiment with pushing a little paper or cotton wool inside to diffuse the display so it looks more grungy but it'll pretty much do as it stands.

Buy cheap buy twice

A few months ago I bought a couple of WTV020SD sound playback modules.

From messing around with the Spookytron I knew the Arduino wasn't really any good at dealing with audio. It consumes a lot of resources both in terms of CPU and pins/timers to access an SD card and then play a WAV file back.

Pretty much on spec I bought these modules thinking they'd 'be handy' as they were very cheap.

I have now wasted several hours spread over a few weeks trying to get them to do anything at all. Even in standalone mode where they work like a basic button driven MP3 player they are simply inert. I've tried everything I can think of including buying some small capacity SD cards as they are apparently fussy about large ones.

So now I'm going to order an equivalent branded item. I've kind of left it too long to return these, but I'm more annoyed at the loss of my time than the money.

Size matters

My next project after the time machine will be considerably smaller and I'm starting to throw a few bits together for it.

This seems like a good opportunity to try out some of the Arduino variants beyond the Uno/Mega. This is a Nano which is pretty much just an Uno in a smaller package.

The plan is to solder it to a piece of stripboard then solder all the connections to that. If it didn't have the headers attached I'd solder the wires straight to it. It's for a permanent installation rather than a bit of quick prototyping.

The project involves interfacing a few things together and connecting them to a PC so I've also bought a USB hub with flying leads which I've stripped down so I can solder straight on to it. There won't be room for a tangle of USB connectors.

Servo board beginnings

I've got seven servos in the time machine and this is too many to control directly from the Arduino with all the other stuff I'm using PWM for.

So I've bought a servo controller chip which connects serially and does all the work instead.

Tonight I started roughly wiring up a board for this, but kind of ran out of steam as I was getting very tired. Over the weekend I really need to get this finished off as it's one of the main things stopping me fixing the dials to body of the machine and finishing up that section.

the board will be based close to the servos to cut down the number of wires I need to connect back with, I should be able to reduce it from 21 to 4.


Having fixed the short I got the flux capacitor and the various other PWM lighting circuits working. This was a good step forwards as so far I've been playing with it all in isolation rather than connected to the Arduino.


Tonight I wired up the MOSFET board and Arduino to the 12V supply in the time machine so I can start making it 'do stuff' independently from a computer.

With the the flux capacitor connected it looks awesome, exactly how I wanted it.

Unfortunately I seem to have managed to create a short in two of the three channels in use on the MOSFET board so they're always on.

I spent a little while last week coding a flux capacitor style quickening chase sequence with some LEDs on a breadboard ready for this moment. Sadly the short means it just kind of flickers a bit.

I've kind of run out of time this evening and have to be up early tomorrow so I'll come back to this later.

The static picture however does kind of show how it's coming along.

Bobba's little brothers

I need to run various things from the Arduino, many of them LEDs I would like to control with PWM.

This is OK straight from the board if you want one or two to mess about with. However I've got loads of LEDs to drive and the RGB strip I'm using for ambient lighting needs 12V.

So I've made myself up a little 10 way N-channel power MOSFET board to drive all this stuff. There's not a massive amount of power going through any of this, but it would be more than the Arduino should really source and well, I need 12V anyway not 5V.

Since adding the flux capacitor to the design I'm a short a couple of channels for all the stuff so need to order some more MOSFETs, but this gives me a great start as I can do most of the LED lighting.

I'm still not 100% sure if I'll use addressable RGB LEDs to light the dials at the back of the machine or just have them lit up white on a single channel. Doing this would make it simpler as I'd be using one method to control things. Also the addressable LEDs are bulky and have short leads between each one so don't actually reach from one dial to the next without doubling them up.

Every time machine ought to have one

You can't have a time machine without a flux capacitor can you. I'd already cut a suitable hole to house it and tonight I quickly threw one together.
I am threatening to turn it into an infinity mirror so I've used a plastic mirror tile as the background.

It's not an attempt at an exact copy of the BttF flux capacitor. I just want it to be recognisable.

The holes are so I can make it light up.
I was lucky to have these bits of acrylic rod kicking around in the cellar attached to an old printer stand. They work really nicely.
The final thing roughly held in place. At some point I have to struggle with the mirror film on the top glass. We've already tried it once and it came out a right mess.


It had been on my mind for some time that if I am making props that rely on Arduinos, files on SD cards and so on I really need some way to tinker with these things in the field.

I don't own a laptop and while you can program Arduinos from Android using a quite nice app called ArduinoDroid, it's a bit fiddly and only supports some boards. Most notably it doesn't support the Mega, which I wish to use.
Buying a shiny new laptop for this would have been a gross extravagance but I was beginning to justify picking up a scruffy secondhand one to myself.

Then a friend asked me to dispose of an old Acer Aspire One netbook she had dropped a year or two ago, smashing the screen. She had finally pulled the interesting files from it using an external monitor I loaned her and wanted me to zap the disk.
Not only did it need a screen, their dog had chewed up the PSU so it was kind of ready for the dump.

However I had  a PSU that suited in my work drawer and these Acers are basic but solidly functional so I wiped the disk, installed Ubuntu and used it with an external monitor for a couple of days to make sure it was passably stable.

Then it was off to eBay for a cheap replacement LCD panel.

Tonight I've fitted this and it was far easier than expected. All I needed was one jeweller's crosshead screwdriver, some care, patience and the foresight to lay it all out on the kitchen table on trays. The only casualty was a tiny bit of plastic clip from the screen bezel. On an old, pre-dropped netbook like this it might have been broken before I opened the case anyway.

Oddly, the Wifi didn't work with the screen broken but came on the first time I fired it up with the new screen. I can only think the screen was shorting out and grounding something that it shouldn't.

So I've now got a tool to work on things out and about for almost free, even if it did suck a little of my time. What I must not do is get tempted into a cycle of upgrades/accessories for this, it would be trivially easy to waste money on a small SSD, more memory, case and so on.

More brass

Working on the Time Machine dials I've wondered for some time how I'll hold the glass on.

I did plan to put brass strip across the front but it was looking a bit ungainly. Instead I cut the strip up and used it to press on the small lip round the bottom of each one.
I'm quite pleased with how this has come out and it holds the glass really solidly.

Currently these are screwed down, but I think ideally I ought to drills some holes in the wood and use M3 bolts.

Dial production line

 With seven servo driven dials to make and time to finish the machine disappearing rapidly, tonight I did a bit of production line bodgery.

As per usual I've done it mostly by eye and stuck it with hot glue but it seems to have come out pretty well.
The dial faces are still the blank trial ones I did so I can just get on with the bulk of the work. Doing a nice design for the faces will have to wait.

If I get time tomorrow I'll make up the brass strips which will hold the glasses in place.

Then I'll have to wire up the servo controller and get it working with the Arduino. There are simply too many things I need PWM for so I'm using a separate dedicated servo controller chip.

First though I must finish building this bit of the machine before I start coding.

Safety first

With the time machine I'm making a pretty fiddly complicated prop with loads of stuff to go wrong.

I'm being really careful wiring it all up and sealing it in place so that things can't come loose or get snagged easily.

However there's a real possibility that it will go wrong and die or just malfunction badly. Equally the batteries may simply not last as long as I'd like.

So I've built in a connector for an external 12V PSU and a 'safe mode' switch. Flick it one way and everything is powered up via the Arduino and driven by the controls/sensors. Flick it the other way and it just supplies power to the plasma globe and under desk lighting using the dumb controller the LED strips came with.

This is already working so if I can't make the progress I would like over the next few weeks we'll have a prop that does something.

Time machine teaser #3

I plan to have a load of dials along the back of the machine, which will be built using servos and a bit of fakery.

Servos are easy to control and cheap as chips. Building a set of dials around them might take a while but will make it more unusual looking than a set of obviously modern moving coil meters.

I've got one vintage ammeter but they're pricey and obviously a 'set' is harder to come by.

Here I've quickly mocked up one dial so I can see how it looks. The 'hand' will most likely have a fine wire pointer on the final version and I'll make more effort with the face.

Pot handles

Having used a bunch of door handles as some of the time machine controls I need a way to read them.

So the obvious answer is mate the square shaft that you ordinarily use with a potentiometer.

I cut the shafts in half and did a bit of very rough and ready drilling so they go straight on the splines of the pot. This was a slightly loose fit so I used a little glue.

Now with these done and mounted I've wired up all the inputs on the top of the desk, apart from the touch switches. Which are very much optional.

Tomorrow I'll connect this all up to the prototyping shield. Ideally I'd like to solder it, but right now I'm going to leave the leads long and just poke them in the headers. It'll be a little flaky but at least allow me to try things out.

At the moment I have about twenty inputs to work with, there are a few more to come including a foot treadle, the handle we bought in Paris and if I can find somewhere to mount them some ultrasonic distance sounders.

Outputs are a little more sparse at the moment, right now the only thing ready to connect up is the vintage ammeter which you can see in the middle of the top.

Time machine teaser #2

Having done a couple more coats of the varnish I'm pretty happy that the finish on the top now looks as intended. So I've screwed the various bits I've got back onto it.

It's not all brass.

Time machine teaser #1

For a LARP we're doing later this year I'm making a 'time machine' of the Victorian inventor variety. There will be brass, lots of brass.

The base for this is a £10 Laiva desk from IKEA, which we've screwed lots of brass stuff to and made various holes in. Now the layout is pretty much finalised I'm going to paint it up a bit so it's mahogany coloured. Here it is with the first couple of coats on.

Very soon I will start connecting all the stuff up to an Arduino I've bought for this job. There will be many buttons and knobs connected then various flashing lights and moving bits so it 'does stuff' in response. If things work out it should make the TARDIS noise and have a homage to the Flux Capacitor in the top right.

Ideally I should have used a nice piece of Victorian furniture for this, but I don't want to ruin something. Also if I mess up a Laiva I can just go and buy another one.

Hurrah for China

Yesterday I ordered an RFID reader/writer. I don't know what I'm going to do with it but it feels like the sort of thing I could have fun with.

It's implemented as a serial device that you send commands to so pretty generic and can be connected with all manner of things including a PC if you feel like it.

This was after wading through the eBay shop of a Chinese supplier with lots of little interesting bits and pieces. There is a lot of really cheap <£5 tat on eBay from these sellers but if you go up in price a bit it looks like there is some really nice kit available.

I was away at a Laser-Tag LARP over the weekend and I loaned them Spookytron for the event. It got used to make a 'haunted house' at the site and worked really well for this. Mik had a few suggestions for improvements which I'll feed into another version of it, but this won't happen for a while.

Right now I've a time machine to build.

Spookytron 1.1

Having used this at a couple of LARPs it's worked OK but wasn't particularly loud.

Great for making spooky noises in the woods but you had to be pretty nearby to hear it.

So I've upgraded to a bigger amp and this should really help.

It also now reads the directories from the SD card nicely and you can swap between directories with the 'mode' pot.

Now I'm going to lend it out to a friend so he can spook his players with it. Hope it stays working. ;-)

Spookytron 1.0 parts list

I've been asked what I built the Spookytron with so here's a parts list. Lots of it came from eBay because I'm lazy.
In total this cost me about £70 before you consider things I had kicking around or got for free like the CAT5 sockets or speakers.

So a fun project but not one that's overly good value. If I refresh it I'll use an Arduino Mega for the extra pins and memory. The Wave Shield eats memory leaving you even less than usual to play with.

Spookytron 1.0

At short notice I decided to go to a sci-fi LARP and the organiser asked if I could sort out some 'spooky noises'. I got a bit overenthusiastic and came up with Spookytron 1.0.

This is an Arduino Uno clone, a generic prototyping shield, one of Adafruit's Wave shields, an opto-isolated relay board and little 1W amp stuffed in a biscuit tin.

This connects over CAT5 cable to four old PC speakers with generic HC-SR501 PIR sensors often sold for DIY projects stuffed in the front.

If a person approaches one of the speakers, sound starts playing on one of the others.

If a person approaches the playing speaker it stops and holds off playing on any of them for thirty seconds.

If it hasn't played for a while it just plays something every now and again.

The SD card is loaded up with spooky sounds. One of the knobs selects between several groups of sounds, one sets the time between random play and the last is the volume.

I took it along to the LARP this weekend and despite it being pretty much thrown together it worked perfectly. We strung about 250m of cable round the site and it played spooky noises overnight through a thunderstorm.

The only things that didn't go quite right were that we put it an area the players didn't approach much so it mostly went off randomly and it wasn't quite loud enough.

The former isn't something easy to deal with and frankly that's what 'random play' is for. The latter will be dealt with before it gets wheeled out again.

I'd also like to improve how it reads the SD card as I'm currently hardcoding the filenames in the Arduino source. Ideally I'd want it to read directories and build the list itself. I was in a hurry when I was writing the code and I wanted it to 'just work'.

If I get really enthusiastic I'll expand it to use eight speakers but I'd need to replace the Uno with a Mega to get the extra lines in/out.