How to make an oscilloscope out of an old monitor
So I'm sitting in my room, looking at the random stacks of computer hardware I have in my loft: a couple of NeXTstations, a few Apple IIs, half a dozen Macs from the 90s...and as I'm looking I come upon this old IBM monitor that I saved from destruction several years ago. Somehow, I decided that I'd try and turn it into an oscilloscope.
Here's the monitor, an IBM PCJr color monitor from 1985:

To start off, here is how a CRT display basically works:

A gun at the back of the screen shoots out a beam of electrons towards a fluorescent screen, and two electromagnetic deflector coils move the beam up and down (usually at 60Hz), and side to side (usually at 31500 Hz), whilst changing the intensity of the electron gun (or guns, with a color screen), to form an image on the screen.
So here is the back all opened up on our IBM monitor.

Deflector Coil Assembly
There are four wires leading up to two deflector coils in the neck of the CRT. In this case, I found by trial and error that the red and blue control the horizontal sync, while yellow and brown control the vertical sync. (Try disconnecting one wires while it's running. If you get a vertical line, you probably just cut one of the horizontal signal wires, and vice-versa.)
Oscilloscope wiring diagram, vertical configuration

After some trial and error, I figured that by keeping the vertical sync connected, and replacing the horizontal sync with an audio signal from an amplifier, you get a nice little oscilloscope:

Vertical Configuration (1 height = 60Hz)
I tried it reverse too; replacing the vertical sync with the audio input and keeping the horizontal sync connected to the monitor supply, but this didn't work as well. I ended up with a horizontal line that would just move up and down with the amplitude of the music.

Horizontal Configuration (1 width = 31.5KHz)
This is because the horizontal coil is moving the beam left-to-right at 31.5kHz, as opposed to 60Hz with the vertical configuration. To remedy this, I tried powering the horizontal coil with the vertical 60Hz supply, but things started producing smoke, so I promptly stopped. I also tried rotating the coil to deflect horizontally instead, but it was epoxied on pretty well. Thus, I settled on the vertical configuration.

60Hz sine wave on the vertical oscilloscope
I also looked up the pinout of the PCJr proprietary monitor connector so I could change the colors. This is achieved by shorting the intensity pin to the color that you want.

PCjr display connector
Here are some shots of the oscilloscope that I took in a dark room. (Sine waves at 60, 120 and 240Hz)



And there you have it. It's not a replacement for an actual calibrated scope by any means, but it certainly looks cool when you drive it with music.
Below is video of the oscilloscope in action. A few observations:
- The beam appears as a dotted-line. This is due to the CRT blanking periods.
- The scope is much more responsive to bass sounds. It appears that the properties of this particular coil cause it to act as a low-pass filter.
- The waveform will sometimes drift upwards/downwards. I think this represents the phase difference between the input signal and the 60Hz vertical sync.
which would be pretty cool!
(of course you'd have to use VNC to control the computer, not the monitor :) )
Course, you could just ssh into the box if it's a linux box, and
echo "hello, dave" | festival --tts
Mark, I enjoyed seeing the oscilloscope in action today. It was good work, and it seemed you had fun doing it. Good luck with the particle accelerator...
now, that is cool
one thing that I am thinking of is something like a little box that you could connect to a vga monitor. the box could have an oscillator to connect to the v-sync pin, and an audio input to connect to the h-sync. you could even have 3 potentiometers connecting the r,g, and b pins to .7v (or 1v, i cant remember), to get various colors.
but would this actually work? I dont think so, because a vga monitor's built in smarts would probably freak it out, what with the constantly changing h-sync. but it would be cool.
question: do I have to use an old-fashioned cga monitor in order to do this? or can I use a vga monitor, and hack it apart similar to what you have done here?
I dont have anything except for more modern vga monitors, but would really like to try this out!
I don't think this will work. I believe most of the VGA monitors today have circuitry to ignore strange h-sync and v-sync signals.
You could make a circuit to provide dummy 60Hz/31.5kHz clocks to the monitor, and put voltages on the color pins, to produce a solid colored screen. Then, you could clip the wires going to the coils and run those to an amplifier to make it a scope.
The CGA monitor was easy because when you fired it up, it had a blank screen by default. It must generate it's own h-sync and v-sync without the assistance of a computer.
You should:
1. completely remove the board under the crt.
2: Unsolder the flyback and connect it to a transistor to make that it generates the EAT(20-40kv) and 210 v for the anodes.Power it up with 12V
3: Connect the transformer pins in the corrects pins of the crt(DON'T USE A TESTER IT WILL BLOW UP WITH 20KV or WITH 200V high freq.)
4:Connect the h sync to a variable sawtooth oscillator(timebase)
5:Connect the vsync to a variable gain bf amplifier
well, I'm sure that I could rouse one up from somewhere.
Although, in all honesty, I'm fairly sure that I've never seen a cga monitor before now (I may have, although I wouldn't have recognized it).
i just made one with a little portable tv and a guitar battery powered amp it works great!!!!! its a jwin tv and a epiphone ep-1 amp an the amp has the speaker cabinet detached, so its just the part with the nobs, and it is taped to the top of the tv.
Hello,
I was very interested in the oscillator you made with a tv set and a guitar amp. I have a gorgeous portable tv that I would like to connect to my (stereo) amp to play oscillator images along with the music that is coming out of the amp.
Is what the owner of this blog did with a CRT monitor doable with a TV?
If not, would you tell me how you did what you did?
Thanks so much.
You could probably do it somehow, but not in the same way. Take a look at all the connectors you could use, and then look up how they all work online - that'll give you a starting point at least. The way these old monitors work has a lot to do with why this is doable without too much modification - on a modern monitor (even a standard VGA CRT) you would have to provide a dummy signal since the monitor tries to use the signal to detect its refresh rate and resolution - or figure out when to turn off.
So I guess the answer is that yes, it's possible to do, but it may be more involved depending on what sort of connections the TV uses. If it has an analog hook-up (the "yellow cable" you might use for game consoles or something) then I'd bet that would be the easiest to do this to.
I built somethings similar to this, but there was a varistor on the board that started smoking when I disconnected the deflection coils. The issue I believe was that the deflectors were in some way important to another part of the circuit. I think the solution is to remove the original deflector coils, but leave them attached, and then slip on a deflector coil from another monitor. That way the monitor circuit will still work while allowing me to play with the deflectors... Also, it is possible to run one channel of audio to vertical and one to horizontal deflector.
turn your scanning coils around 45 degrees and youll get your display working properly
rab
how would you make the crt monitor work without a visual signal ( mine goes to sleep if its not connected.)
The CGA monitor I used for this project only needed a few pins shorted out on the monitor cable to make the screen solid with a particular color. (Shorting the intensity pin with a color pin.) Upon recieving an intensity signal, the CGA monitor itself appears to drive the display with hsync and vsync signals. This is possible with a fixed frequency monitor (one that only runs at a single, set resolution.)
This would not work on the typical VGA monitor, most of which are designed to run off several different possible resolutions (multisync). This kind of monitor expects to be driven by a proper signal before anything will display on the screen. There are a few ways around this: You could make a circuit to supply phony 640x480 sync signals at the VGA port input, then set a color pin to 1V (or whatever it is for max intensity). Alternatively, you could drive the coil directly with a sawtooth wave, while also finding a way to activate the electron gun directly.
If I ever get around to doing this with a VGA multisync monitor, I'll be sure to post my results here.
Hey I have been trying to do exactly what you did that that ancient CRT with an '82 color TV. What i have done is; disconnect all wires connecting main circuit board to coils, soldered both horizontal wires coming from the coil to where the vertical wires were coming from the circuit board, and connected the wires from the vertical coil to a sound source.
http://i157.photobucket.com/albums/t54/rediculus08/FW2NDNJG0KQK5BKMEDIUM.jpg
However since I believe that the reason It did not work for me is because I am using a color TV. So I thought that maybe your project with a color CRT might be somewhat similar to a color TV. My question; what exactly did you do?!
The fact that it's a TV shouldn't make any difference for this project. As long as the electron guns fire some sort of image (like static), the coils can be manipulated to fire with sound instead of the sawtooth raster scan.
What does your configuration produce on the screen?
My configuration was different from the one you drew. The vertical sync was left completely intact, while the horizontal coil was detached and driven with an audio signal. I tried doing what you did (driving the horizontal coil with the vertical sync), in the hopes of producing a more traditional-looking horizontal oscilloscope, but promptly stopped after something started smoking within the circuitry!
It ends up that the two coils (H and V) are too different to switch the signals like that. They were designed to have an appropriate frequency response for their respective sync signals (horizontal sync is in the tens of kilohertz, while vertical sync is around 60 hertz), and thus are built with different impedances. Simple illustration of this: when driving the horizontal coil at 20000Hz vs 200Hz at the same volume (amplitude), the 20000Hz wave shows up much smaller on the scope. In this case, the properties of the coil where such that it was acting as a low pass filter.
Just out of curiosity, do you think you could measure the resistances of the two coils (H and V) with a multimeter? Also, visually inspect them: do the horizontal vs vertical windings have noticeably different turns / lengths / thicknesses? (It's been a while since I've torn one of these apart, it might not be easy to see.)
The configuration I mentioned above produces a blank screen, nothing. I also do believe that for a TV, the oscilloscope should be horizontal due to there being more hertz and the waves would be more visible as a traditional oscilloscope would display them (I'm not too sure on this, just something I read somewhere).
Also, I do not have a multimeter lying around so at the moment I am unable to measure the resistance of the coils. And no the H and V coils did not have noticeably different turns / lengths / thicknesses. Here are a couple pictures:
http://i157.photobucket.com/albums/t54/rediculus08/DSC00013.jpg
http://i157.photobucket.com/albums/t54/rediculus08/DSC00012.jpg
So maybe I should also try leaving the vertical sync intact while driving the horizontal sync with an audio signal? When you tried the traditional-looking horizontal oscilloscope, did you notice any visuals before it started smoking?