When I bought my house there was a wireless push-button for the doorbell, this was mounted on top of an old door intercom system. Because the plate is metal, it acted as a shield making the wireless door bell not work very well. Next to this was an plastic house number that was cracked all over from the sun UV light.
Time to make something new, and off course I made it a bit overkill.
The old situation:
I drew a design for an stainless steel plate. This plate contains the house number, a hole for the push button, and 4 mounting holes. This design I had lasered by RoboCNC.
A piece of smoked PMMA plastic will be mounted behind this plate. It will be illuminated by a LED board:
This board has 3 modes:
White light
Adjustable RGB
RGB fade
The lights are switched on / off with an build in twilight switch.
It is not convenient to change the mode on the build in board. Therefore I made a “remote controll” that will be mounted in the house fuse box on a din rail.
When I etch my own circuit boards, sometimes it happens that there is a very small short on the board. This short is difficult to find. To burn this short away it is handy to have a power supply that can deliver lot of current. My heaviest lab supply however can only supply 6A. Therefore I made this project:
This project is a 5V power supply with an adjustable current limit up to 30A. This power can not be outputted continuously. To make the 5V I used an old ATX PC power supply, these can supply a lot of current on the 5V line.
I have put 12 high power transistors on the secondary side of the power supply to regulate the voltage/current. The transistors are mounted to actively cooled heat sinks. 2 extra power transistors are added to drive the other transistors.
The heat sinks are from 2 old Pentium 2 processors. On top I put a few fans.
I found a few LCD voltmeters in my junk supply. These I have used in this project as volt and ampere meter.
Test layout:
Mounted:
With the switch the output can be enabled. The switch has 3 positions:
-Neutral position = OFF
-Spring return position = output on as long as the switch is activated
-Latch position = output permanent on
The 2 panel meters need their own power supply (isolated voltage), this I solved by adding a small transformer followed by 2 7805 regulators.
This part is the drive circuit of the transistors. This also include the current measurement:
Circuit:
The big pile of resistors are used because the power in the resistors at these currents is very high. I could have used a lower resistance value, but I did not have that in stock. From these resistors I had a partial reel (SMD).
In the pictures in the previous update a cd case is used as a safety glass in front of the drill. This is rather thin, therefore I decided to exchange it to 4mm plexiglas. This new plexiglas is hinged by the cover of an old tape deck.
I needed a acrylic plate that is 90cm long and is bend 90° over the ling side. For this I requested some quotes from various suppliers. They all came in at about €100 to do this. Therefore I decided to invest the €100 in a DIY bending jig, this way I can also use it in future projects.
After some searching in my scrap-parts bin I came across a heater resistor from a clothes dryer.
I cut of the ceramic isolators and mounted the resistor wire between 2 screws. Around this I build the bending part. This consists of a fixed board and a board on hinges.
I added some aluminium foil that is covered with a layer of kapton isolaton tape to protect the wood against the heat
An extra board is screwed on top to fix the acrylic sheet.
The device can be used like this, however it gets too hot and may discolour/warp the plastic. Therefore I added an PID controller to control the temperature of the wire. In the center I stuck a thermocouple sensor to the resistance wire.
The messy inside of the controller box:
A videoclip of the bender with a polycarbonate plate:
(I also heated from the top with a heat gun, this to get it hot faster. This may not be necessary, however I have not tested yet without this.
The result:
When copying this design please not that the resistor wire is connected to mains, and therefore is not safe. To use this more safe a isolation transformer is recommended.
Some time ago I bought a lot of Weller soldering irons at an auction for cheap. Between these where a few DSX 80 de-soldering irons.
I also had an old Weller IG101 solder station (bought earlyer from another hobbyist), unfortunately I had no iron for this one and the DSX 80 could not be connected to this one. Time to convert the IG101 to newer standards.
IG101 station:
To use this station a few alterations where required. Most important is that it is 24V input, I want to change this to 230V. Also I did not like the connection for the vacuum to be on the side of the unit, so I moved it to the front (same as the modern units). The vacuum is created by a venturi valve, the air output of this valve is still on the side, since the enclosure was too small for another bend.
First I closed the old holes for the air connections and added a cut-out for a 230V power connector. The cover will later be painted making these “fixes” invisible.
I placed the valve on the bottom of the enclosure. On top of the valve a metal plate to mount the toroid transformer (2x12V 100VA)
Next the front panel was changed to incorporate the vacuum connection, Nokia 5110 LCD and buttons.
The old electronics where replaced with a new board driven by a PIC microcontroller.
In a lot of electronics I bought there was a small variac. Unfortunately I could not find a datasheet, but after some experimentation I found out it was for 12V AC. Laying on the shelve it was quite useless, so I decided to make an low voltage AC power supply with it.
To determine the max input voltage, I measured the zero load current and put it with the input voltage in a graph:
From this graph I concluded that as long I stay below 15V saturation is not a problem. Therefore I have chosen to use it on 12V.
I build it in enclosure together with an 12V halogen transformer. Also I readout based on that of my Safety transformer V2 & Safety transformer V3, the version in this project however is adapted to measure lower voltages, so the resolution increases. Also a power supply for the meters is added.
My new workshop is in my basement. For the hand washing / board cleaning “corner” this has the disadvantage that it is below the sewage level. In order to drain the water it needs to be pumped up. I had a few pumps from old washing machines laying around and decided to do some experimentation with this.
To make it easier for the pump (and to spare my back) I made a frame underneath the cabinets to raise the whole thing by 20cm. This way the water needs to be lifted less.
Normally there is a siphon after the drain in the sink, this part I have taken out and brought the pipe directly to the pump.
At the pump side the pipe goes down a bit in order to make the check valve work better (Picture further down on the page). Also because of this the sensor is not mounted at the lowest point. This makes it work better since the pump does not clear all the water from the pipes. The water that does stay in the pipe may false trigger otherwise.
However this means that the pump needs to operate longer when the sensor signal go’s off. Therefore I made a timer box. This was made with a Siemens LOGO The time is settable by a rotary switch. This switch is connected with a few inputs of the LOGO.
The pump:
The timer box:
The program in the LOGO!:
To detect the water in the pipe I used a E2K-L26MC1 capacitive sensor from Omron. This sensor can be clamped around the pipe and does not come into contact with the water.
The sensor cheapest supplier for this sensor, I found on AliExpress.com: – Omron Level Sensor E2K-L26MC1
This sensor has an NPN output, while the LOGO is only compatible with PNP output sensors. This I solved by adding a 10K pull-up resistor between out and +24V. After this the input needs to be inverted in the LOGO to work correctly (B17 in the program above).
For a project I need to program a few microcontrollers in the same circuit. This meant I needed to plug the programmer around on the board a lot. This got annoying very fast. Therefore I decided to make a switch box. In my junk pile I found an old switch of a, parallel port switch. This has 4 positions and a lot of contacts. For the ICSP I only need 3. However in some circuits the supply voltage is not common. Hence, I chose to also switch the power supply connections. For the connections to the circuit boards I used DIN connectors, for the simple reason I have lots of these.
I needed to cut a lot of wire to specific lengths. At first I did this manually with a different colour wire as a reference, but after a few wires it gets annoying/boring quite fast. So, time to automatise it. Off-course there are machine’s that can be bought to do this, but they are rather expensive, therefore I decided to try and make one myself.
The first version I build did not work very good unfortunately. It was made on top of an old power supply enclosure and made with thin wood. This made it all a bit too flimsy.
For a test version I made a video:
After this attempt I bought a 3D printer (not specificity for this project), and made a second version with 3d printed parts mounted on a thick base plate.
To measure the cable length and feed the wire at the same time I made an assembly based around a stepper motor. The wire is fed between 2 rubber rolls. One is scrapped from an old VHS recorder and the other one from an old HP printer.
After the rollers the wire is fed trough a tube to the cutter., this way the wire is fed correctly trough every time.
To cut the wire I used a cheap side cutter and a geared DC motor to operate it.
With the old version the wire spool was just laying on the floor. This made the wire twisted when it got to the rollers. I bought a big bearing at a radio flea market in Eksel, this bearing I used to make a spool holder.
To driver everything I made a microcontroller board that contains a H bridge to drive DC motors, also it contains a stepstick stepper motor driver. For putting in the parameters I reused the front panel from my PC clock project since this was no longer in use.
When I wrote the firmware it turned out that it was not very convenient to set the wire lengths etc with the 5 buttons. Therefore I replaced them with a keypad.
The bipolar stepper motor used above was not strong enough to move the wire reliable. I therefore replaced it by another one I had laying around, this one was a unipolar model. This meant that I could no longer use the stepstick, so I replaced it by a unipolar stepper driver I made before.
I also made an attempt to make a automatic spool winder, to putt the wire in a nice bundle again. However this did not work as expected, and after a few attempts I decided to cancel this part.
The final result can be seen here:
In the last couple of years this project has cut a few kilometres of wire for me.
The circuit diagram I can not share because of the undocumented changes. The code can be found here.
I have a Steinel Gluematic 3002 hot glue gun. This glue-gun itself I’m fairly happy about, but the mains cable is much too stiff. This causes it to tip over from the force of the wire.
Therefore I decided to replace the cable with something more flexible. I came across some silicone rubber wiring. This is very flexible, but also is heat resistant. This means it will not get damaged if I accidentally touch it with the hot nozzle.
After opening the glue gun it turned out to be very easy to replace the cable. The strain relieve was not molded to the cable and the wires to the heating element where just crimped to the other wire.
Replaced:
This modification makes it much easier to work with.
We use cookies to ensure that we give you the best experience on our website. If you continue to use this site we will assume that you are happy with it.
Dutch 
