Easy DC?

Well, six months after the fact, I believe I may have located the source of the pretty disastrous issues I had at the last Pi Wars. I actually managed to complete just three of the seven events so it was all a bit heartbreaking tbh.

I had severe issues using a PS3 controller and I fully believed that to be the issue and therefore have looked at moving away from this to another controller type for next time round. The Rock Candy PS3 controllers come highly recommended by Brian Corteil (aka @CannonFodder) who has used them to great effect, in fact winning – twice!

So, you may have seen FRED-209, my new Pi powered Nerf Tank. This uses same ZeroBorg motor controller and I had it linked to an original PS3 controller. I’ve tested extensively in the week and then this Saturday I took it to Cotswold Jam and bam… the controller fails. Niet… Zip… Diddly. Can’t get it to connect, then if it does it drops after a few seconds.

This is *identical* to the behaviour X-Bot exhibited at Pi Wars. So much discussion ensues as to the cause. I know of several other folks who also had issues with these controllers last time round. However, Brian C. lent me a Rock Candy one to test and this also showed exact same behaviour. The RC ones are 2.4ghz and the PS3 ones are Bluetooth so still wasn’t sure if it was radio interference of some kind. Another good guess was noisy DC motors causing problems.

But… I think we have a winner. This DC-DC converter board is a damn liar…


I recently recommended this type of thing having seen similar at Pi Wars last time. I now conclusively retract that recommendation! I used a similar board last time , without a read out, and I believe this suffers from the same issue.

Note the difference in alleged and actual voltage….


This difference is not enough to shut the Pi down, but it is enough to mean power hungry peripherals, say a bluetooth dongle for example, are just enough to make the power dip enough that the USB experiences issues and that kills the controller.

If this were the only problem then simply setting it higher to start off with would solve the problem. We did crank it up to 5.18v on the multimeter which may be getting a little hot for the Pi, but it then ran flawlessly for the rest of the day at Cotswold Jam with loads of people giving FRED-209 a good hammering.

I do though believe there is good reason to suspect that the setting on the board drifts over time. Whether this is vibration from the pot moving the micro screw that you use to set the output voltage or something else, it does seem to change. This is regardless of the state of the batteries being used.

So, I am now hunting for a DC-DC board that sticks out a guaranteed 5v that can be relied one. Core Robotics have one in the pipeline but it’s not available yet. I’d be interested to hear from other teams who used these or similar boards or PS3 controllers last time and also experienced issues.


Until next time….

Read More

Say hello to FRED-209… (You have 20 seconds to comply.)


This one’s been a while in the making, I’ve had a whole summer of fun putting it together. Tbh I can’t remember what first inspired the idea for this. I’ve been wondering about Nerf guns for a while (another amazing thing that wasn’t around when I was a youngster) My main thought was ‘can you use a servo to trigger it?’

Turns out you can :)

First experiments….

Flushed with initial success, things got out of hand remarkably quickly….


Whilst a fun idea, it obviously wasn’t going to fly that way so I decided I’d see if I could design a ‘Nerf Tank’ from the ground up. I’ve use the ZeroBorg controller again from Pi Borg and a PS3 controller (despite one dying on me at PiWars!) I chose this as the base as I have working controller code and I wanted to focus on the actual 3D design and build this time so I could complete in time for the next Cotswold Raspberry Jam.

The Nerf in the pic above is the Elite Stryfe, about £15 from various places. Or – your local car booty for a whole lot less! ..


I pretty rapidly disassembled it to get a good understanding of the innards. Turns out it’s two flywheels that actual push the dart out and a fairly simple push mechanism to give it a shove. Hmmm… two motors – and a servo. This could work!


I then got OpenSCAD out again, still my favourite 3D design tool, I love its overall simplicity and the fact you can do fairly complex shapes in just a few lines of code (without having to learn what a zillion buttons do) I drew out a basic design for the chassis and I made larger versions of the same motor mounts I made for the PiWars bot.

This is the original chassis I designed, combined with the gorgeous chunky wheels I found on eBay. Sadly the motors turned out not to have enough grunt to move the thing so a rapid and somewhat disappointing switch was made to the Rover 5 chassis you can see in the final version (which was also an eBay find) Also in the below picture is the design for the tilt mechanism which would eventually be controlled by two of the shoulder buttons on the PS3 controller.


Here’s the completed barrel and mechanism box, you can see the servo mounted underneath which attaches to a nice simple pusher. Printing the lid was fairly tricky as it and the box are the largest single pieces I’ve printed at 24cm long. The lid also required a lot of support material which was a bit of a bugger to remove cleanly as it covered such a large surface area.


Servo pusher…..


And flinger bit….


I had to tweak a couple of dimensions on the box to make sure everything lined up nicely but it went together really quite quickly in the end.

Testing time!!

You’ll notice this version is black. And really quite poorly printed. After two crappy results I finally worked out something pretty important. Cheap PLA isn’t worth the saving. I spent many hours trying to work out why the results were so poor when previously the printer had been excellent. Turns out it wasn’t calibration, or heat settings, or bed temperature – just naff filament. It works ok on smaller items but the moment you try and print something larger you’ll run into all sorts of problems with warping, lifting and delamination.

I therefore binned the black, and splashed out on some shiny orange (flouro red was the as described colour but hmm…) I also had to redesign the bottom plate to fit onto the Rover 5 chassis. It doesn’t look as good as it did on the chunky wheels but hey, it works.


I’m planing to take this to Cotswold Jam so thought it may be a good idea to give the kids something to aim at (apart from each other!) So – evil alien hordes it is. I’ve also rigged up a Pimoroni Blinkt, the lovely little 8 LED board. This works as a visual countdown, starting with all green LEDs and then one going from green to red every 15 seconds. The plan is to mark out an arena and then the player has two minutes to knock down as many targets as possible and then return to base before time runs out.


Doing a little target practice with the finished build…

(Apologies for rough quality of video – it was downsampled during upload to youtube for some reason.)

I’ll put the STL files for the chassis and the Nerf cannon on Thingiverse when I get some time later this week, I’ll also upload the controller code to GitHub.

I’ve got some future plans for this chap too. Adding a camera and web interface controller seems like the logical next step – and guaranteed to cause remote mayhem! Further to that, I’ve started experimenting recently with OpenCV so am very interested in adding a level of autonomy / target detection. Hopefully FRED-209 will be slightly better behaved than his namesake!

Thanks for reading – pop along to Cotswold Jam on 23rd September if you want to have a play – there’ll be prizes for the highest scores!

Until next time…..

Read More

Weapons of mass distraction.

While I’ve had the last few months to freak out about having no motors and therefore essentially no bot, I’ve had plenty of time to try and distract myself and design and build a range of add ons to cope with some fairly stiff challenges in this year’s competition.

I built a 4-way sensor rig from standard ultrasonic sensors. The wiring is pretty um… intense, but it does work and reads distance in all directions. I’ll be using this for the straight line speed test and the maze.

The wiring maze begins…..


All wired up….


I’ve also got a mount for the line sensors sorted and the code working nicely for that (fingers crossed this works on the day!). I’ve also got a simple holder for the golf challenge, although I suspect there may be far sneakier / trickier hazards than this will cope with! :/ The one on the left is the clamp for Pi Noon… Looking forward to that one.


Skittles was great fun last year, and my ‘ball cannon’ worked well enough, sadly my driving let me down and I knocked the ball away from the bot and fluffed a shot. The ‘collect the ball’ component of the challenge has forced me to come up with something to avoid this completely this time round.


The simple arm / servo arrangement should prevent the ball rolling away, and the spring in the tube has enough kick to give the ball a good whack!

Armed and (slightly) dangerous…


I’ve got the bot driving properly and the mecanum wheels are performing really well – such a relief at this point. I’m now (slowly) working on the autonomous challenges and rapidly running out of time. Going to be a frantic dash to the finish line. I reckon I’ll get there :-)

I’ve got a bot that can compete and even with all the difficulties I faced, I’ve learnt so much again through the the build and it’s been really fantastic to be involved in Pi Wars again. I love the social camaraderie, particularly on Twitter. Very much looking forward to seeing old faces and new next weekend.

I’ll be there for both days, do come and say hello. I’ll be the one with the panicked look cos his bot just * shot out the door / took out a small child / knee-capped Eben Upton (*delete as applicable)

See you all next weekend!

Until then…

Read More

Motor Woes – Part Deux

It never rains…..

I have not as yet managed to assemble a complete set of four matching motors in the six months I’ve been in preparation for this competition. Unreal.

And, with 10 days to go, I managed to strip two of the gear boxes on the motors I have. I took them apart and discovered that the gears in my ‘metal gear motors’ were in fact…  plastic. Seriously? – After how long it took me to find them?

Something’s definitely not right…

Ok… DefCon 1 level panic time. Some great suggestions on Twitter (thanks folks!) and some fleabay scouring actually brought these beauties to my attention thankfully pretty quickly…


Not cheap, but sooooo pleased I found these, 24hrs ago I was pretty much dead in the water.

These motors are 12v and I have been running at 9v. I’ve stepped it up to 10.5v which is the maximum voltage the Zeroborg controller I am using can handle. They are nice and torque-y and are rated at 500rpm, I reckon I am getting 430rpm (ish) at the voltage I am using.

Although I built the motor mounts to be removable it still required a complete strip down of the chassis to remove and replace all four motors (again).


I’ve 3D printed a fake AA battery and run a wire through it so I can use an 8 cell holder with only 7 cells in it to give me the required 10.5v

Works a treat!…


I’ve also removed and replaced the rollers for the ninetiteenth time. This time to add some thread lock. The wheels do put some fairly significant forces through the rollers and they’ve had a tendancy to spit bolts on the odd occasion so I took the opportunity to tighten everything up.


Anyhoo. Another 48hrs flies by, and I have rarely been so grateful to be back where I was two days ago. Onwards – I still have a ridiculous amount to do. At least now I have four matching motors!

I’ll hopefully get time for one more post where I’ll do a quick overview of my wild and frankly heath-robinsonian attachments for some of the challenges.

Until then….


Read More

My Achilles Wheel…

As I’ve previously mentioned my last Pi Wars bot was rudimentary to say the least. This time I wanted something that would push my skills further and also result in a better designed / built bot. There’s nothing like jumping in at the deep end so when I got my hands on the 3D printer last year, I had an idea. A mad, stupid, almost impossibly, ridiculously complex idea but hey – this is Pi Wars, right? I’ve kept this largely under my hat during the development of these as I honestly wasn’t sure I could get them to work…



So, I saw PiBorg and their mecanum wheeled robot at a couple of events and pretty much fell in love. This combined with witnessing the performance of the omni-directional bots last time was enough to convince me to attempt to do something similar. I looked around and found prices for mecanum wheels to be more than a little eye-watering and therefore decided to see if I could 3D print some. It became pretty clear, pretty quickly, to be honest that the full design is waaaaay beyond my capabilities so I adapted an original set I found on Thingiverse which were just so freakin’ cool.

The first frame appears on the printer…..


They were designed for Lego motors originally so I adapted the hubs to work with Pololu 4mm hubs and gear motors. I also reconfigured the roller attachment to work with M3 bolts, rather than Lego axles.

Roller up…. Roller up.

Each frame takes around 7hrs to print at 0.2mm resolution – fortunately all four printed first time with no errors.

And now the fun really begins…. 4 Wheels. 8 rollers per wheel, 2 bolts per roller.

I’d originally intended to print in NinjaFlex – a really cool rubbery 3D printing filament but sadly it turned out my printer couldn’t use this so I had to get creative.

Printing the basic rollers….


(Each roller takes 19 minutes to print)

I can print in PLA or ABS, both of which are basically solid plastic and therefore completely unsuitable to be used as rollers as they have no grip at all. So – a covering was required. To quote myself and many other Pi Wars contestants – ‘How hard can that be?’

As it turns out, very. Annoyingly, swearingly difficult actually. I have now tried and tested seven types of covering. That’s all 32 rollers removed, coated, replaced, tested. Often twice – just to be certain.

Hanging ’em out to dry….


The list of what didn’t work includes…

  1. Balloons
  2. Heatshrink – thought this was guaranteed but failed completely!
  3. Rubber Cement – cracking idea from Graeme C. – sadly it peeled right off.
  4. Latex – dipping each roller many, many times to build up the layers. Very weird stuff – and ripped straight off.
  5. Grip Tape – stuff skateboarders use. Useless, wouldn’t adhere to rollers at all.
  6. PlastiDip – rubber coating stuff. Took ages, cost a fortune. And ripped off, again.

It can be seen from the above that coating and grippiness wasn’t an issue, durability was. So, as I was about to throw in the towel, Mark from Astro Designs came up with idea number seven…

7. Self Amalgamating Tape – aka plumbers tape.

And….. we have a winner!


Here’re are the rollers wrapped with tape and ready for testing. It had become a proper love hate relationship with these damn wheels by this point in the process.

Thankfully, I think we’re looking good. Whilst the bot is ridiculously ‘slidey’ the wheels do actually work.

I conservatively estimate I have spent 100+ hours on the wheels alone so I am thrilled that I have a bot with ‘proper’ mecanum wheels. Whether this offers any advantage in the competition remains to be seen!


The tape offers enough grip to get the bot to move and it’s durable enough not to tear straight off. It’s a complete riot to try and drive the thing however. It may yet be christened ‘Uncontrollabot’ !

It has of course made autonomous driving and the associated challenges much, much harder than they may well have been. But hey, we’ll cross that bridge when we get to it.

In about two weeks time :-O

Until next time….

Read More

Lightning does strike twice

Just a quick update this time. Got a lot on at the moment!

If you were around for the last Pi Wars you may remember I fried a Pi for the first time ever whilst mucking about with relays for the skittle ball release mechanism. It was sad. We had a minute’s silence. RIP RPi.

Weeeeell, it would appear I doggone went and did it again. Nooooooo!! This time it was whilst trying to hook up the Adafruit servo board (which I’ve used successfully several times in the past.)

There was smoke this time too. Proper thick, acrid, really not good smoke type of smoke. Rapidly followed by lots of frantic wire pulling.

Was undoubtedly a classic ID10T error. A quick aide memoir for myself and others:

9v > 5v > 3v3

This is important. It goes wrong quickly if you mix them up. Very quickly.

I was in all honesty surprised when the Pi actually booted up again. Sadly though, a quick check of the i2c deelybob (that’s a technical term I think) revealed that all was not well…


I’m not sure therefore if it’s the i2c driver on the Pi that’s gone, or whether its the servo board. In reality, I’ve not had time to check, I swapped out the Pi for a fresh one and have limited myself to using the hardware PWM on GPIO 18 of the Pi. I’m only running one servo that holds an arm to hold the ball in place (hopefully) for the golf and skittles events.

I had another moment earlier this week when, on reviewing the challenges on the piwars.org site that I realised that there had been a small, but rather flippin important change to the skittles challenge – someone put a turn in the course! I’d not seen the important ‘competitors must collect the ball and then…’ bit. The wonderful @GeekyTim ‘fessed up on Twitter when challenged and admitted that it was indeed he who had added a little something extra. This should make it a lot more interesting and requires some thought in actually controlling the ball to even get it as far as going in the direction of the skittles.

I’ve still got a long way to go, have been manically designing and printing a wild and wacky range of attachments to serve in the array of challenges. More of these next time. The bot is being assembled and reassembled repeatedly as I test, fit and retest different configurations. I’ve got most of the manual challenges sorted now and am working on the challenges that require some form of autonomy. I *may* have made these an awful lot more difficult than they reasonably should be due to a critical design decision I took very early on in the process. I’ll let you know.

So – onward and PiWar’d. Ermm…. did I say there’s only two weeks to go? Gotta dash – I’ll catch up with you all later.

Until next time… (more…)

Read More


Finally feels like things are starting to come together. Literally. Since my last bot was made from two cobbled together project boxes, any amount of design is going to be a step forward. I’ve continued to experiment with OpenScad and am really loving it. There’s a really good cheat sheet here that gives you pretty much everything to design anything of any shape. It’s really a very powerful FREE tool and tbh makes much more sense to me than the AutoCad type programs. I’d also previously tried SketchUp and found it to be the same type of thing with a million incomprehensible buttons and a very steep learning curve. For my purposes OpenScad suits perfectly.

So, I’d made the motor mounts (for non-existent motors at the time) and wanted to build a fairly simple chassis to mount it all on. To the 3D printer, Batman! I put together a very basic shape, with holes for the motor mounts, and then, in trying to make it look at least a little less basic, began experimenting with chopping bits out of it. This can bo done really easily with the difference() function in OpenScad that takes one shape away from another.

I then saw fairly quickly where I wanted to go with this and….


Behold – X-Bot is born! :-)

I am printing in PLA with 20% fill, which gives plenty of rigidity without taking a age to print anything. Got to admit that having the 3D printer is an absolute gem. Being able to conceive an idea, design it and hold it in your hand a very short time later is still pretty science fiction to me!

The motor mounts went on really nicely. I’m really pleased with these considering my motors were held on with hot glue last time!


And once all four were mounted…


I started to lay out the electronics to get an idea for the spacing etc…


I also attached the wheels for the first time – Yay! For the first time it actually looks recognisably like a Pi Wars bot.


Was also REALLY pleased to see my calculations were indeed correct and it does fit on an A4 sheet of paper – phew.

So, the basic chassis and control mechanisms are complete. I *still* have motor nightmares to confront. The seller who owes me a back order is yet to come up with the goods so I may yet have to find another plan.

I’m going to focus on the manual challenges next as planning / coding for the autonomous challenges is a non-starter with un-matched motors. I’ve got an idea for the golf challenge and am going to rework last year’s skittles cannon so hopefully can sort those two fairly quickly. I’m designing swap-able mods for each of the challenges, I spent too long having to reconfigure the bot during the day last time so want to avoid this as much as possible.

Can’t believe there is sooooo little time left!

Next time – Weapons of Mass Distraction.

Until then….



Read More

Motor woes… And tantrum throws.

We do this because we love it.

We do this because we love it.

We do this because we love it.

(Keep repeating the mantra)

It’s not that bad, but some parts of the process can make you swear a little. Additionally, it’s 48hrs to the Pi Party, I’ve got loads of prep left to do – and have had lovely tonsillitis for 3 days. Seriously hoping the antibiotics kick in soon!

So – four motors right, just four little motors. No problems. So, I started ordering motors back in November and after two hopeless ebay sellers took TEN WEEKS to fail to deliver anything to me at all I was getting a little sweaty by the time the new year crawled round. So, third time’s a charm (as the saying goes) and I found another supplier who actually delivered the goods in about a week.


Last year’s motors at the front – new ones behind. Ooooh – look at the shiny shiny!

I was sooooo happy to get my hands on these then a few things started to dawn on me. The motors are 12v. I am using a ZeroBorg from PiBorg, the very neat little 4-channel controller they first launched as a Kickstarter campaign. Turns out the ZeroBorg can only take a maximum of 10.4v, so I am actually running it at 9v. Bang goes 25% of the power going to the motors immediately.

In addition to this, the motors are rated as ‘300rpm’ (when running at 12v) So I expected around 200-225rpm at 9v. This I would have been happy with. The motors I used last year were 130rpm but really torque-y. These new ones actually turned out to run SLOWER?!! Whoever gave these their speed rating needs a good talking to IMHO!

Back to square one.

Another year, another supplier (Ireland this time)….

‘Hi, you have 300rpm, 6v motors?’


‘And they’re definitely 300rpm?’

‘Oh, yes’

‘And they’re definitely 6v?’

‘Oh, definitely sir’

Fantastic. So, I order and a week goes by. I am juuuust getting antsy when a parcel lands on my door step.

Containing three motors.

And a note ‘Sorry – we’ve only got 3 of those atm, we’ll send you the other on back order when we get it – possibly by March.’


Are. You. Kidding me.

(Oh – it’s now March, and I am still waiting.)

So, I find my FIFTH supplier, in the US this time, who has the exact same motors that I have three of, but at a premium price plus US / UK shipping costs. I therefore spent a lot more than I should to get a matching forth motor. In fairness to the seller, he shipped immediately and it arrived within a week.

And guess what? – The SAME motor, from the SAME manufacturer, with the SAME label actually has a 6mm shaft not a 4mm as the others do. And it runs significantly slower than the other three. Utterly frustrating, but can’t really blame the seller for that one, although QC at this particular manufacturer obviously leaves a lot to be desired!

So, with 6 weeks to the competition, I am yet to find a fully matching set of motors. I have compensated in the code to allow for this, but it does mean slowing the other three down to match which isn’t really great – I certainly won’t be winning any speed runs I feel!

One a positive note, I’ve designed my own motor mounts for standard 25mm gearmotors, I’ll put these on Thingiverse so they’re available. Very pleased with the way they turned out…



I’ve also got the Pi Zero / ZeroBorg wired up and have started tweaking the code. I’ve also got a PS3 controller set up for manual challenges. This is a significant step up in usability and complexity from last time round. Think I’m also going to need a lot more driving practice!



Next time…. Get it together, man. – Time to get chassis-tastic and try and actually build a bot!

Until then…





Read More


Well, I think I can be forgiven for the title, hey – I’m a little excited, ok?

The last competition was undoubtedly one of the highlights of 2015 and indeed of my Raspberry Pi experiences (and I have a fair few of those now!) So I was thrilled when in October (REALLY?!) I, along with the other teams, received the email from Mike and Tim telling us we had been successful and had been selected to take part in the 2017 competition. There were many more applicants than places so I am extremely pleased to have got in. However, the next week or so were taken up writing a PhD. proposal and in the latter part of October I found I had been successful. I was not due to start until February but my supervisor wanted me to start ASAP so life sort of went out the window at that point. I am currently living away from home and am based in Milton Keynes from Monday to Friday. This has made planning and particularly building a bot more of a challenge than I had expected.

Last year my entry, sHazBot was built almost entirely from recycled junk – something I am very proud of. She came a respectable 4th overall and had the brazen cheek to walk (well, roll) off with first place in the obstacle course challenge. I was even lucky enough to have my wife capture it on film…

sHazBot certainly performed admirably but with significant short comings. Whilst the motors I chose were extremely torque-y they were also pretty snail like and the bot certainly was not going to be setting any speed records. The control system was also fairly rudimentary with simple on / off controls for the motors. This was ok for some events, but really not suited to the events where precise, directional, control was required.

I therefore came up with a hit-list of what I wanted to achieve this time round:

  • Faster. Much faster
  • Proportional Control
  • Encoders on wheels
  • Better chassis design
  • More aesthetic. (I came LAST for aesthetics last time round!)

From winner… to the bits box.

It happens to us all eventually. We go from the top of our game to the rest home in a far quicker time frame than we can possibly imagine…..


In this case, most of the parts have been donated to other projects – including this one! I took the motors out and did some tests before deciding that I was going to invest in some new ones. I also wanted a better designed chassis – well, to be frank, one that had been designed at all would be a good start.

If you follow me on Twitter, you’ll be aware of my latest acquisition and new found inspiration, I finally sold off a lot of old gear and scraped the funds together to invest in a 3D Printer. And a damn fine one too. I have seen too many people spend too much time fixing, calibrating and building 3D printers. While I’ll concede that this is a fine way to learn, I wanted a 3D printer to actually print things. So, I did a lot of research, settled on the exact model I wanted and set up alerts on eBay and Gumtree. I only had to wait a week or two until what I wanted popped up at the right price so after a quick trip to Leicester I returned home the proud owner of a (wait for it..) Wanhao Duplicator D5S Mini.


All I can say is very impressed indeed. The version I bought was a 2015 model which did not have a heated bed. The 2016 model does, which is kind of an admission that it needs one. It does. I experienced warping on large, square prints and tried pretty much every combination of glue / tape / spray etc and was still having problems. I therefore bought a third party heated bed which is considerably cheaper than the Wanhao version and have had perfect prints since. I’ve had enormous amounts of fun mucking around learned an awful lot and have printed many useless, and a few useful, things.

I really wanted to design the full chassis for the bot and 3D print it. I was much impressed by Tito, Hitchen Hackspace’s fully 3D printed bot from Pi Wars 2015. I’ve been teaching myself to use OpenSCAD which is a language based 3D modelling application. Really powerful but simple to get started and to be able to do enough with easily.

This week I have been mostly designing motor mounts and brackets.



I was very impressed that the printer managed to print the complete piece with no supports necessary. I’ve got loads more design work to do on the chassis, I’ve got the motors and control board / controller to sort out and that will then give me the basic bot to work up from. I’ve got most of the sensors I am planning to use but need to get the main bot functioning before I can start to add complexity.

Long way to go, and the weeks are flying by. Did I mention I’m moving house too between now and the event?

Next time……. Motor woes and tantrum throws!

Until then….




Read More

Halloween Fun – A Spooky Ghoul-in-a-Box!

So it’s that time of year again. The UK seems to have become more upbeat about Halloween in the last few years, certainly a difference from my youth when trick or treating was distinctly frowned upon! We always get lots of little horrors dropping by so we have a big bowl of treats waiting – along with a little surprise.

You may have seen the ‘Pumpkin Pi’ I made last year, this was a complete hoot and worked very successfully.

This year I thought I’d try something different – and a little more animated! I like the ultrasonic sensor that the Pumpkin Pi uses, it’s really easy to set up and there are loads of resources online – Python is my favourite flavour and it’s really easy to set up simple distance measuring with the sensor which can lead to all sorts of Hocus Pocus.

The initial thought was, I want something hidden that can jump out of a box with a ghastly scream. Achieving that took a fun afternoon of hackery and a lot of bits and pieces I had lying around.

So, to build a motion-sensing-ghoul-in-a-box I used the following recipe:

1 garden fence trellis

1 bicycle inner tube

Lots of scrap wood

Lots of glue, staples, tape and wishful thinking.

1 Raspberry Pi

1 Ultrasonic Sensor

1 mask (35p from Tesco!)

1 box

Making the base…



I saw the trellis in the back of the shed and knew that it would be perfect. Getting enough spring for it to ‘jump’ quickly and effectively was going to be tricky. I looked at elastic bands and the few springs I have lying around and none was going to have enough oomph. I’d seen some really powerful catapults made using tyre inner tubes so I found an old one with a puncture and re-purposed this.


I stapled the sliced up tube to the trellis and then attached them to the frame like this..


Then added a small screw as a latch to hold it down under tension (you can’t see the screw in this picture – but you can see it holds it nicely!


Then it was time to go inside and put together the electronics. This was quite straightforward, I set up the sensor and the Python code to measure the distance. I then took a relay board I designed for sHazBot, my PiWars robot and used this to turn a motor backwards and forwards rapidly to act as a ‘trigger’ Releasing something held under load can be tricky but this method works very well.


I 3D printed a little cam to push against the trellis and release the spring…


Time to put it all in the box – added a speaker to play some horrible screams, and a battery so it can all run headless (how appropriate) and it all fitted in nicely.


Finally I added a *really scary* (?!) mask that cost the grand sum of 35p! I also added LEDs to the eyes for extra effect.

Spooky Ghoul-in-a-box – primed and ready!


I think the end result works really well, think I’ll be leaving this out on our front drive on Monday to welcome any callers who may happen to pass by!

The video is a bit dark and best viewed in a dark room. But hey, it’s a horror movie! It’s also a LOT louder when you’re standing next to it (I turned it up to 11)

(No ghouls were harmed in the making of this video.)

Have a terrifyingly happy Halloween everyone!

Until next time..





Read More