Electromechanical "magic eye" display
Original idea and implementation by sv3ora



Magic eye indicators always fascinated people. But a magic eye tube may not always be available when you want it, plus some time in the future good condition eye tubes will be rare. Not always that but the high voltages involved, make this a project unsuitable for the inexperienced. Also, magic eye tubes come in predefined sizes/shapes. Therefore I decided to reproduce a magic eye in electromechanical means. I first saw this idea on this website but it included lots of mechanical parts and certainly it was not a simple project. Therefore I experimented on different ideas of how to make the mechanical construction as easy as possible. I also created my completely own schematic, the first version of which I present below. Later on in this article, you will find more elaborate updates of the magic eye display.

My initial circuit uses only a few discrete components. The two 2n2907 transistors, form a simple comparator and the two 2n2222 transistors a schmitt trigger and led driver. In my circuit, the range of voltage which the display shows can be adjusted (display deflection) from about 100mV to about 5V. There is also a preset minimum control, so that the display stays on (but at minimum deflection) even on the minimum voltage level.

Because I wanted to use this indicator as a useful general purpose measurement display device, I included two imputs. The one at the left, accepts DC (100mV-5V) and the one at the right accepts audio or RF AC (0Hz-30MHz tested ok, but it can go much higher), which is detected and converted to DC to drive the comparator. Because the indicator is to be used in different applications requiring different responce times of the display, I included four electrolytic capacitors at the input, that can be switched in a decade capacitor configuration (one at a time or in combination), which slow down the response of the display accordingly.

The emectromechanical display is very fast, much faster than an analogue needle meter, so it is capable of displaying the peak power in SSB transmitters. It is also very accurate, the "beam" is straight line and not bended like in many magic eye tubes. Thus you can do actual measurements, by placing a scale on the perimeter of the "beam" of the indicator.

As far as the mechanical construction concern, the first version of my electromechanical magic eye indicator, has been built using a CPU motor. These motors have 3 wires, two of them are the DC to the motor and the third, is a feedback from the magnetic sensor inside the motor, which is used to control the speed of the motor. These motors can be almost noiseless when rotating, especially if you cut the fins, like I did for this experiment. Here is a short animated gif of my first version in operation.


In this version, I greatly reduced the mechanical complexity, due to the CPU motor that contains all the parts inside. However, this can only lit from the front now, which raises another problem, glare. A LED must point to the front surface of the motor, and shine a target (yellow line) which in turn reflects the light. But to be successful, we must ensure a high contrast between the black area and the yellow line and glare is our enemy. I did different experiments, but the best results for this version of the indicator, were obtained by using a transparent lamination film and laminating a paper below it, where it was painted yellow. Then I painted the film top surface with thick black permanent marker. The led must point to the film from it's sides at an angle and not directly from it's top. This reduced glare to the minimum and gave a high contrast to the indicator as you see in the picture below.

The result was good but not perfect. If you notice the picture above, the permanent marker (which dries really fast) have been kept long onto the film surface and dried. The second time the marker tip passed over the film to make a second layer, distorted the finish of the previously dried marker. These little distortions reduce contrast in a fast spinning disc.


The picture above shows my second experiment. This time I used a larger diameter CPU motor and I did not laminate any paper. Instead I used the transparent film alone and I painted it at it's external surface with the same permanent marker. The way I painted it was, I clipped the film onto the motor shaft with a small magnet and I let the motor spin at low speed. Then I put the marker onto the surface of the film and like a gramophone, I scan through the film surface with the marker. This cave a much better finish. there are still some distortions but these are much much smaller now. Use the thickest marker you have to minimise these distortions. This time, I also cut a small piece from a luminescent green-yellow sticker I have found and I glued it onto the transparent film surface, to form the indicator.


The animated gif above shows my second version. The shadow at the top is caused by the large height magnet I used, but this was just for demonstration purposes. The usage of the magnet allows for easier alignment of the film. As a light source, I found white LEDs better, but this depends on the luninescent indicator material. If you use a UV LED, the glare is much reduced and the indicator shines bright. But for eye safety reasons I did not use one. Because of the usage of a visible LED, you have to make sure that if you enclose it into a box, the white LED must not shine out of the box to your eyes, but only to the film surface.

The second version worked great and if you really do not want to mess up with any mechanical parts, this is the way to go. However, there are still things that can be improved if one wants a better result. The glare, although minimized, can still be seen by looking at the eye from an angle. The permanent marker drawn disc is something that might not be very durable and it is affected by dirt and hands fingerprints. Also, the usage of the CPU motor has a property that can be thought as a disadvantage or not, depended on what you are trying to do. These motors reset twice per revolution, so the eye pattern is repeated giving two beams instead of one. Now, many magic eye tubes use this two-beam pattern, so this can be considered as a more realistic effect. However, this reduces the screen resolution to 180 degrees instead of 360 degrees. Not only that, but the magnetic sensor inside these motors causes the resolution to decrease even more and in fact the usable pattern angle is a little more than 100 degrees.


UPDATE 19 Nov 2021

In this version, I have used a totally different way to build the eye. I have used back lighting, which offers the best brightness and contrast. I have also increased significantly the display range of the eye, which is now almost full circle.



To achieve these features, I used a small DC motor and attached to it, a piece of polyamide material, cut on the lathe. At the side of this material, a small ball magnet is attached, which upon rotation of the motor, engages a reed switch nearby. In the schematic at the beginning of this page, the reed switch replaces the internal magnetic sensor of the CPU fan motor. The switch is thus closed once per revolution.



Two high brightness LEDs are placed in parallel connection, between the polyamide material and the motor, thus lighting the molyamide from below. Their light is refracted very well and evenly in the polyamide material. At the top of the polyamide material, a disc of laminated paper (painted black) has been cut in shape as shown in the pictures and clued to the polyamide. To achieve even brightness a special cut has been done in this disc. The circuit voltage has been increased to 8v and it is stepped down for the motor, using a set of 68 ohm power resistors (22 ohm shown in the picture), connected in potential divider configuration. To be able to solder the LEDs directly to the motor body and keep the body to ground, I changed the LED driver circuit by connecting the final transistor collector to the VCC and the emitter through a 27 ohm resistor to the LEDs.



The circuit works amazingly well. The brightness of the eye is great and can be seen very well even on direct light. The mechanical construction is still kept rigid yet not many moving parts are required.


UPDATE 13 Aug 2022

In this version, I used an optical feedback DC motor out of an inkjet printer and I changed the circuit to work with this motor. The motor is shown below.



The new schematic that works with such motors, is shown below. In this schematic I have also added several features. Apart from the positive input, there is also a negative input. The negative input can be set with the trimmer so that the eye can reach to maximum deflection at a certain negative voltage between about minus 2-3v up to more than minus 20v. This is the range of most magic eye tubes. Therefore, this negative input can be connected directly to the grid input of the tube it replaces, and so it is a drop-in replacement for the tube without any modifications to the radio!



The next change, is the addition of the automatic mosfet switch and the timed/free jumper. This jumper enables or disables the motor sync pulses/switch. Because of this you can now apply an audio AC signal to the imputs and notice the different nice display patterns as you change the audio frequency. Also, because of this, you can now apply arbitrary pulses with correct width and timing to the aye display and create a decatron effect!

I have also changed the LED circuit and now the LEDs are grounded, which makes heatsinking easier. I have also used more LEDs in parallel, high brightness green SMD LEDs. An interesting feature of the optical feedback system, is that you can now create different eye patterns, by just marking the feedback disc holes you like. Here is an example of a pattern I created by marking all holes onto the motor feedback disc with a silver permanent marker, apart from 4 holes which I left transparent. The pattern is split into four sections, each of them filling up, as the input voltage rises. The more sections of course the less the range of the eye display.




The picture below in contrast, shows the pattern with just one hole in the disc left transparent.




The circuit was implemented onto a single prototype board with the SMD LEDs soldered at the solder side, just below the polyamide and the laminated paper disc.




Back to main site