Intro
CinemaScope was a breakthrough lens technology that compressed widescreen images onto regular film (and decompresed them in theaters) and helped movies compete with television. Here is the story:
https://www.youtube.com/watch?v=Bve8wGAPhIg&ab_channel=AtomicAgePictures
This project is not about a lens, I just wanted to explain how the name was similar.
This project is about displaying raster video on an oscilloscope - it takes the cinema prefix and abbreviates oscilloscope to arrive at Cinema-Scope.
I have been itching to try displaying video on my oscilloscope, but the scope is not really designed to show raster video.
It can show X-Y inputs which allow vector graphics to be displayed, but it is missing a Z input to modulate the intensity of a raster scan.
There are a few scopes out there that do have a Z input, but mine does not.
Nevertheless I think it may be possible to display black-and-white video with a little extra circuitry.
This is the story of that journey....
This image is from some graphic art I did 25 years ago.
Circuit Description
To convert a raster video into something a scope can display requires creating a raster trace and modulating it. A raster trace sweeps a trace across the screen for each line of pixels and then moves down one line at a time. To sweep a trace across the screen requires an increasing voltage on the "X" input this voltage needs to instantly reset at the end of every line - which means the X input needs to be a sawtooth waveform at the line frequency - which is about 15734 lines per second for NTSC video. Similarly the Y input needs to decrease its voltage for each line of pixels and quickly reset to maximum voltage at the end of each field, which implies an inverted sawtooth waveform at about 59.94 Hz in the case of NTSC.
My scope has the ability to display X-Y inputs but unfortunately it does not have a "Z" input to modulate the signal intensity.
However it should still be possible to externally modulate the signal on and off to provide a black-and-white display.
The circuit wouldn't actually turn the signal off, just switch it to zero volts when the display should be black.
Here is one of the circuits I came up with:
The PCB looks like this:
The PCBs cost $5 / 10 cards, and arrived 3 days after ordering.
This circuit has a composite video input feeding an LM1881 which extracts horizontal sync pulses and vertical sync pulses.
These pulses are used to reset op-amp integrators (using analog switches) to create sawtooth waveforms. The "horizontal" sawtooth is a ramp that increases with time where the voltage corresponds to a position in the X axis of the scope - when it is in X-Y mode: The op-amps used in this project have a 10 MHz bandwidth, because I am using one of them as a comparator at pixel frequency, but it could probably use an even higher bandwidth op-amp or a proper comparator.
The frequency of the sawtooth in this case is 15.714 KHz. The flat section is due to the horizontal sync pulse duration.
The "vertical" (inverted) sawtooth has a decreasing voltage with time because its voltage corresponds to he vertical position of the trace and the raster scan of the video expects to start at the top of the screen which is the highest voltage.
This waveform has a frequency of 59.94 Hz.
When these sawtooth signals drive an X-Y display they should create a "white" screen as the signal scans back and forth over the whole screen.
The circuit also has a comparator to turn the voltage of one of the waveforms to zero volts (using an analog switch) whenever the video signal exceeds its threshold - which is adjustable using a trimpot.
My circuit has an issue with displaying the raster as I envisioned it. There is some noise in the signals that cause the pixels to randomly be displaced a bit, which makes the image fuzzy.
The scope also creates a problem because it is digital. When the circuit is trying to turn off the trace, it switches the voltage to zero in a few nanoseconds. On an analog scope this wouldn't be visible because the trace brightness is proportional to the time spent at a voltage, so turning down the brightness would make fast signals disappear. On my digital scope it actually will display some of these fast signals (because they are actually there) as a sort of waterfall smear.
Here is an example of the 2 issues:
To display a composite video signal I used a Raspberry Pi and connected it to a small composite display at the same time as the Cinema-Scope circuit.
I spent quite a bit of time trying to minimize these issues - without much success.
I tried switching the horizontal signal to zero instead of the vertical - with similar results.
Running the "USB" 5 volt supply from a well regulated supply (instead of a USB power module) made a significant improvement.
I also tried a bunch of other methods to reduce noise, but it will need more radical measures.
The circuit originally had a third (video) output to modulate intensity, which is useless on my scope so I hacked it to provide a sync for the scope, when modulating the vertical sawtooth.
Here is what the CCA looks like after I started hacking it to try different fixes:
Test Setup
The 5 inch TFT LCD (800x480) was surprisingly legible - all the small text generated by the Raspberry Pi was readable.
Test Setup Video With Example Video Content
I don't think I will be using this system to view video content, but it was an interesting exercise.
I will continue to ponder how to improve the circuit, but in the mean time, here is a video showing some of the graphics I created for the project as well as how the circuit and scope fare when displaying video:
Here is an image showing results of filtering the vertical sawtooth to a half millivolt of noise and switching the horizontal sawtooth at pixel frequencies:
I did not show build details of the 3D printed stand for the little composite monitor since it wasn't central to the project, but if you want more info on any aspects of the project, or if you have any suggestions on how to improve system performance, let me know in the comments below.
Here are some images to show that this scope can do other colors (about 6 colors including rainbow and fire multi-colors):
Adjusting the scope can provide a more accurate image, but it is grainier (this is with filtered vertical sawtooth).....
At some point I will try to find an analog scope with Z input to see if the circuit works better with an instrument like that.
Relevant Links:
Pretty Cool Raster Text Display on an AD3
Project14 | Photography: We're Giving Away Raspberry Pi HQ Camera Kits for Projects that Use Them!
Top Comments