Follow by Email

Tuesday, 22 December 2015

Beach Huts

These parts

plus some more colourful parts

Once welded together with a soldering iron and painted

create a set of beach huts. Of course you have to make sure that the colour combinations are acceptable with some sorting:

More to follow when I have the full set of parts...

Commodore SR4148R Calculator

In order to get a charger for a Commodore S60 that I recently bought, I purchased an SR4148R which came with the same model of charger as the S60 uses. The S60 had old leaky NiCds in it which I replaced, and a faulty VFD display with bad ghosting that I fixed with a pull down resistor.

Anyway, the charger for the SR4148R worked fine, so I have that for the S60. It didn't turn on the SR4148 though. From the outside it looks like this:

From the inside it was apparent why it didn't work on the charger:

There were no batteries and the charger socket wasn't wired to the main board. The PCBs were covered in battery leakage mess. After a bit more disassembly wwe get this:

Taking the keyboard apart led to a nice surprise. the keys are individually mounted and each one has it's own metal spring. A quality design that simply wouldn't be done these days.

Very unfortunately, you can see that the battery leakage has reached the LED display, as there is green behind the bubbles, on the left.

I cleaned the mess with baking soda and it all cleaned up a bit. I then disassembled the LED display, removing the bubbles to see if I could dissolve the mess off the display. After some cleaning I managed to do this and got a mess free display.

Very very unfortunately, the mess seems to have completely dissolved two of the seven segment digits. They just aren't there any more. Oh well, that isn't going to be repairable at all, I suspect. I will now have to find a broken calculator of the same model, or a replacement display. Interestingly I noticed that the small 0.96 inch OLED displays that I am fiddling with at the moment are an interesting size relative to the LED display:

It looks like two displays would cover the same area as the original display. Adding a PIC micro and some code to demultiplex the drive signals and we could display the seven segments on the OLED display. Hmm, a project for another day, but it probably isn't too tricky.

Monday, 14 December 2015

Printing Station

Having bought another printer (the Fisher) I needed to accomodate the two printers in the workroom. I had already built a plywood container for the Mendel that I used to store the printer and also transport it about before I had a permanent home for it. As I wanted to keep this case in case I wanted to move the Mendel again (and also protect it against dirt when it wasn't in use) I decided to fit it with legs and put the Fisher underneath the Mendel. This gives me a printing station with two printers one on top of each other and an area for filament reels below.

The small coloured filament is being used (I store it in bags when not in use) and is hanging on the frame for some wall mounted IKEA CD holders.

Here's a video of both printers running simultaneously:

The base for the mendel is fixed to the wall using the type of bracket that I print out quite often:

Creating a bracket like this is really quick and easy (in OpenSCAD) and while it is being printed I can do something else...

Small Box

In the olden days when yoiu wanted a small box for a couple of audio jacks attached to a ribbon cable, you'd have to send a pretty large cheque through the post to a mail order firm, then wait, then it arrives and you'd have drilled some holes, and there you go...

These days you print one.

Faberdashery Filament Storage

i have bought quite a lot of filamanet from Faberdashery in the UK. It's very good quality, but it doesn't come on a rell, which can be a pain. I have made reels in the past, but the coil of filament needed to be wound onto the reel, which was time consuming.

So, I made a reel out of cardboard and printed blocks:

The advantage of this reel is that you can unscrew one of the cardboard sides, slip a coil of filament in and screw the side back on. This makes it much easier to fit the coil to a reel.

Once the reel is screwed together I slip it on a filament bar that I fitted to the printing bench.

There's a couple of cardboard reels on the filament bar, the one on the far left is a wooden version I made, which is a lot more robust, but the cardboard ones have been pretty good so far. They also have the advantage that the corrugated cardboard is perfectly formed to act as a filaement end holder, as you can see on the black filament reel in the middle. You can tuck the end of the filament into the cardboard corrugation hole to stop the reel unwinding.

Surface Mount Component Storage
 I've bought some surface mount components as they seem to be the way ahead. They make it easier to make PCBs as you can skip most or all of the process of drilling holes in the PCB, at least if you stick to single sided PCBs. The problem with SM components when compred to the old through hole components is storage. I ordered a bundle of resistors and a bundle of capacitors from ebay as a starter set. They are quite compact, but awkward to store in an orderly fashion.

One common way to store these components is in plastic wallets which are then held in a ring binder. The pockets are designed to hold components, either on a strip like these, or in a small envelope. i need to store strips, and having to delve around in an envelope seems a bit fiddly, so I looked for plastic pockets. Unfortunately they seem to be pretty tricky to buy and are expensive. Some people on the web have tried making their own pockets out of standard A4 plastic ring binder holders. This involves a soldering iron or a hot wire system to melt the plastic and build up smaller pockets. This looks a bit hit and miss and also quite a lot of work. So, I gave up on the pockets and stored the components in a drawer.

Until I saw this in a charity shop this weekend:

It's a (very old) photo album. It is a binder with card pages that have a plastic page on each side. The plastic sticks in a temporary fashion some tacky strips on the cardboard pages.

The idea is that you peel back the plastic and tuck your photos inside, under the clear plastic sheet. The tacky strips then hold the sheet down and stop the photos sliding about.

This is also ideal for holding the strips of surface mount components:

You can get quite a lot of strips on each page, I suspect too many would reduce the number of tacky strips that are available to hold the plastic, but 6 strips is no problem at all.

As the binders were designed to hold photos they have been designed to expand as items are put into the binder, which also means thst they can accomodate quite a few strips of components.

I managed to buy two binder of different exterior design but with the same tacky strip arrangement and I'll definitely be looking for more as I suspect they can be used to store things other than SM components.


I've had an old BT phone for a while now, and recently bought a cheap one in a market. I finally decided to put some time into converting it to stand-alone use, using an arduino. The result is in the video below.

The internals comprise an Arduino Uno and a 'Music Player' shield. There's also a prototyping shield with some custom circuitry on it.

The original dialler mechanism has two GPIO lines on the Uno attached to it, the code uses these to decode the numbers that are dialled. The 'on hook' switch is also attached to a GPIO line, this is used to detect 'on hook' or 'off hook' and put the code into one of two modes.

The bell solenoids have been rewired from their original series configuration and have been attached to a GPIO line each via some power transistors. 

The music shield is wired to the earpiece in the handset, this is used to play various tracks on an SD card inserted in the shield.

The two modes of operation are either 'on hook' which is when the phone delays and rings, and the 'off hook' mode which is when it waits for dialled numbers. 

Dialled numbers are commands to the phone:

112 moves to the next track on the SD card
2DR is a command that delays for D seconds in 'on hook' mode and then rings the bells for R cycles.
3DR is the same as 2DR except that the delay is in minutes not seconds.

The music player shield is frankly rubbish. It is  an MP3 player chip with GPIO lines wired to all the manual controls. There's no detection of the track being played and all you can do is turn it on and off and move to the next track. And adjust the volume. It barely works for the Arduphone. I have a long recording of a dialtone as the first track, so that is what you hear when you pick up the handset after turning the phone on. It does run out after a while, though, and will then move to the next track.

The ringing circuit was tricky as it is driven from a high voltage in the real phone environment. I had to drive a power transistor with four GPIO lines to get enough current to provide a kick to the clanger which would actually ring the bell. This area could be improved.

I think if I get round to doing a second version then I'll have to update the music shield, but this version is quite a lot of fun eveb with it's problems.


Wednesday, 9 December 2015

Microscope Reflector

A few months ago I bought a couple of old microscopes from a junk shop. They were pretty cheap, but the carriages were seized. I managed to free them up and they work nicely, but on one of them the mirror used to reflect light onto the sample was missing. It was a simple disc shaped mount, so I turned one up on the lathe:

The metal is shiny enough for this to be good enough to use without a mirror. i may cut a piece of mirror to fit, or, more likely, I will add some form of LED illumination. Messing about with a mirror is a bit of a pain anyway.

Woodrat Things

I'm still working on a template routing attachment for the wwodrat. The latest part I'm trying to make is a clamp for holding work to be morticed. The latest version of the part is here:

the idea is that this is screwed to the template holder plate and the bolts hold the workpiece. This is 3d printed in PLA, and unfortunately it isn't rigid enough, I need to reprint with a different design.
It takes about 2 hours to print this, so I have to pick a window where I can allow this much time to print the part, so it's taking a while to get this done.

Thursday, 3 December 2015

Shredder Pushstick

We have to cut a fair amount of hedge down and bought a shredder to mash up the cuttings, as otherwise they take up a lot of space. The shredder came with a plastic pushstick for pushing stuff into the shredder. It was a bit delicate, though, and started to tear in half after about 8 hours of shredding. So I made a wooden one:

I've yet to try it out, we'll see how well it works...

Monday, 30 November 2015

The Robot Moves!

The robot has been wired up and some rudimentary code written. It moves. The bumpers are all wired up as well and so it has collision detection.

The video shows the robot randomly moving around the workshop floor. It actually pushes all the debris on the floor to the edges of the area it covers, which is essentially what I hoped it would do. I swept the debris up with a dustpan after it stopped, which was easier than sweeping the whole floor.

Next step is to add a vacuum system so the robot collects debris rather than push it about.

Also, I'm not sure the cheap batteries I bought recently as battery packs are anywhere near the capacities they said they were. Hmm, not too much of a surprise bearing in mind the price.

Saturday, 14 November 2015

Watch Bracelet Spring Success

It looks like the latest iteration of a replacement spring for my Breitling Aerospace has been successful. After the spring died (after only 16 years...) I first tried unravelling a spring I had lying around and used it to make a replacement. It worked OK, but as the original spring had been circular it was difficult to make it a nice shape. The bracelet has a flat cavity for the spring nd it just didn't lie flat and caught on the bracelet as the spring action was exercised.

Attempt 2 (well 2,3 and 4) was a 3d printed part in PLA. It worked well and was perfectly sized. Unfortunately it opnly worked well for 2 days, then it lost all springiness and was useless.

I then bought some stainless steel wire (wire supposed to be used for safety wiring engine parts), but it was too brittle to form into a spring. It only allowed one bend, any attempt to rebend and the wire snapped.

Next attempt was piano wire. I ordered some 0.51mm piano wire and formed it into a flat spring with just a few spring elements. The original spring had a clever design that allowed quite a few elements to fold into a small space. I couldn't form it in such a complex way, so I am trying less elements. This allows the spring to fold into a small space and hence gives more extension when putting the watch on.

Wednesday, 11 November 2015

Spoon Collection

I made some wooden spoons a while back but didn't post photos as they were a present and it was all a bit hush hush. Now I can post some photos, here they are:

Acomplete set of different types and sizes.

Tuesday, 3 November 2015


I'm putting together a robot that will be used to trundle around the house, hopefully picking up dirt and so on. That's the plan anyway.

The design is an octagonal base with two stepper motors as drive wheels. An arduino Mega will control the robot, and there will be sensors for bump-detection and maybe position determination.

The stepper motors are geared down, and the gearbox has a 6mm shaft witha 5mm flat. I couldn't find any wheels that fitted that shaft so I printed some:

I'm not sure how well the elastic band tyres will work, I may have to find some other grippy way to do tyres.

The stepper motors are mounted with printed parts too:

The stepper motor drivers are individual PCBs, I printed mounts for them too:

I'm probably going to mount two battery packs, one for electronics and one for the motor drive. This should keep noise from the motors at a minimum and should also allow the electronics to still function when the motor drive is dead. I could them switch the motors to use the electronics battery pack and limp back to a charging station. For now the robot will flatten it's battery packs and they will have to be removed to be charged externally.
Anyway, more printed parts hold the battery pack:

The arduino Mega has a prototyping shield on it, where I will add extra circuitry and also attach the motor drives and sensor inputs.  It is also mounted with printed parts:

The parts are mounted on a plywood base, this is octagonal and will have bumpers on each edge for collision detection:

Bumpers next, they are a bit tricky as microswitches aren't that straightforward to mount. I've started with a single printed bumper which contains two microswitch mounts and some flexible spring parts. So far I have quite a few iterations of the bumper:

In that photo the first bumpers are on the right hand side and later iterations are to the left.
They still need a bit of work, but they are nearly there. I need to make sure the switches are operated from all directions and everywhere long the length of the bumper.

I also have an ultrasonic sensor that I may mount on a servo so I can measure distances in different directions.

Shutter Hooks
The shutters on the workshop doors have a tensioned spring that is designed to provide an assistance when lifting them. It works well and the effort required isn't too bad, even with the big shutters. At the top of the travel, however, the spring isn't under much tension, and so the shutters have a habit of drooping after they have been lifted. I thought about drilling out some lock slots in the runners, but that would have required completely dismounting and disassembling them, and I don't fancy that. So, off to the 3d printer again, and I made a hook and hinge:

 This is the hinge and hook assembled as it will be used,

and this is the hook in use on the big shutter. The force it needs to exert to hold the shutter up is very small and PLA is easily strong enough.

Watchstrap Spring Saga
The spring that allows the bracelet on my watch has broken into pieces after only 15 years of use. It's a spring that can compress to a very small volume.

My first attempt at a fix was to see if they were on ebay, but no, not really. There are some springs that may work, but they weren't definitely the correct one and were very expensive. So, attempt one at making a spring was an unwound spring bent into roughly the correct shape:

Not pretty, but it sort of worked. It wasn't quite the right length, so there was still a bit of slack.

Next attempt was a few 3d printed springs in PLA:

These are much neater and a better fit as the sizes can be tuned easily and accurately. They work well too, but only for a while. The PLA seems to creep and end up in the compressed state, i.e. sort of lose it's springiness.

I've ordered some piano wire, we'll see how that works out...

Monday, 2 November 2015

Jasper Trials Finished

Well, almost finished, I still have to slide the glass in and hang it on the wall. 

This now has the painted number, the paper labels and pencil numbering. Everything I could see on the original. The colours aren't 100% correct, but that's almost impossible from photos.

Sunday, 25 October 2015


The new shutters for the workshop rooms are now fitted. Fitting wasn't too difficult, but the shutter curtains themselves are pretty heavy.

The large shutter up...

and down...

In the down position the shutters protect the doors from any bits and pieces that may get flung about by machinery (it happens now and again...).

The smaller room has a shiutter too, this is up:

and this is down:

Not much gets flung about in the small room, but they are useful for security.

I had to pack the shutters out by 20mm in order to clear the door handles, that was pretty simple. They take up very little space, the main space taken up is above the doors and that isn't particularly useful space anyway.

Thursday, 15 October 2015


One of the storage heaters has had a broken knob for quite a while now. It's easy to work around as there's two on the heater, and I can swap one for the other when I need to adjust one of the settings that the broken knob is attached to.
Anyway, time to see if I can print a new pair of knobs as a fix.

Old knobs:

The new knobs are in grey and have a similar but simpler design. I needed a pointer mark, so used a sharpie, but I could have put a recessed shape on one end of the cross piece, having thought about it.

This is them in operation, and they do fit and work pretty well:

Wednesday, 14 October 2015

Jasper Trial Trial
Of course, in the process of making a jasper trial tray it is necessary to perform some trials in order to test various ways of making things. So, you end up with several trial pieces for different trials in the tray you are reproducing. And what do you do with these trials?
You build a trial trial tray.

These trial pieces have some of the writing that is on the original pieces (the Wedgwood museum was kind enough to send me a high resolution picture of the tray so I can determine what the text says for all pieces). I trialled different ways to write the text.
The numbers are stamped with number punches from a 5mm set ( I have ordered a 6mm set as these numbers are too small for the full size trials).
The colours are not correct in these pieces, colour is difficult to reproduce accurately, anyway.

The box for the tray is a trial, too, this is what the full size box will look like, stained this colour too.


Woodrat Template Routing

When making chairs, I am going to be making a lot of tenons, so I thought a way to template route on the Woodrat would be a good idea. There is the new Pathfinder, but it seems a bit elaborate for me, mainly due to it having to be able to do a large range of templates.
If you have a 3D printer then it's pretty easy to print a template specifically for each job, then you don't need a template grid with all sorts of different tenons etc.

So, I am attempting to make a template routing setup for the woodrat. A plate of MDF is attached to the workpiece extrusion that moves left and right. A recess is routed in the MDF to allow different templates to be inserted. A pin is used to trace the template and it fits into a pin holder that is attached with slots so there is some front and back adjustment.

This is the bracket at the front of the underside of the router plate. It holds the end of the MDF plate that is nearest the user.

The far end of the MDf plate is held with two brackets that slide in the extrusion slots. They are held with M4 hex head bolts attached with butterfly nuts so they can be tightened easily. This provides the left/right adjustment of the template so it can be aligned with the workpiece.

The template itself sits in a recess routed into the MDF:

There's a couple of tabs on the template which are used to attach the template firmly with a couple of screws. The tenon in that photo was cut with the template fitted.

The pin holder is attached to the router plate and has front/back adjustment.

I need to improve the stiffness of the pin holder and make a proper pin that has less friction than a bolt with a thread on it.

I also need to make a similar system that works on the mortice rail that I made so that I can cut repeatable mortices in the chair legs.
That will need a new pair of brackets and maybe a new MDF plate.