Quote:
Originally Posted by Tube Radio
Off topic, but saw your website and realized that's where I got the info for my Bell & Howell # Model 9560-1 5 Inch Triggered Sweep Oscilloscope Kit
I originally thought of using the display in an old 70's Sun automotive diagnostic machine to display a scope clock given from what I understand the display is basically a vector monitor, but never did much with that idea given issues I have with it.
|
I'm very glad the manual was of use.
Does that Sun have a 7" round P1/green CRT or another kind? In the 70s, I had one like that made in the late 50s with tubes and its bandwidth was only around 50-100KHz. Long before electronic ignitions. I was a kid in the 70s, and tinkered it back to life, sold it to a hot rodder.
RANT ON
As a note and a small caution about scopeclocks, relating to my last post -the part about 'point-plotting' scopeclocks (as opposed to line-and-circle-drawing): kits and other things like instructions abound for point-plotters, and there are more of them now than ever. Those are the cheapest and simplest of scopeclocks and require only software and two D-to-A channels on the processor.
"buy this random small-CPU dongle thing and use this software to make a cheap scopeclock".
Not picking on this one, but it's an example of a troublesome issue:
an ESP32-based wi-fi dongle board was used to make a simple scopeclock, and the
youtube video shows it with an electrostatic scope, and the obvious jitter and inaccuracy of the plotted display is blamed on the
Advance Instruments OS1000A scope.
The scope is called 'outdated' by the author, yet it has a bandwidth of -3dB at 20MHz! Sure it's old, but 'outdated' is a subjective term. In no way should a scopeclock board/device need a 20MHz scope! That's nuts. 1MHz should be enough for a good match between a processor and a thoughtful piece of software. I displayed the
Sparkfun O-Clock point-plotting scopeclock on a
450KHz Tektronix RM 503 XY scope. It also looked great on a
200KHz H/P 120AR. That was a thoughtful scopeclock board and software product. (Its code and making documents are still there too, free. Based on the Dutchtronix O-Clock)
The plotting issues with the ESP-32 based scopeclock shown in the video might be that the processor is running the WiFi radios etc, getting NTP time, and on top of trying to plot all the points for the figures in the the clock. That processor is likely enough for both, but I would badly guess that the scope display code could be made to use less CPU cycles per second - plot less complete scope faces per second, and the issue might go away. Disclaimer: I don't write software so how would I know? - But it's inappropriate to require a >> 20MHz scope for use as a scopeclock. Maybe the scope as-shown has an electronic problem (ripple in supply etc). As an analog scope used in XY mode, I can see that the issue is not inadequate scope bandwidth specifications. The display would be distorted in a different manner if that were the case.
My points:
1.) programmers using fast processors should not disregard the relatively low or slow performance of the common and inexpensive old electrostatic scopes that hobbyists will have available for these experiments.
2.) A consideration should be that the hobbyist will rather buy a 1-3MHz $5 scope and fix it for this purpose, not waste a 100MHz instrument on it, or even a 20Mhz one.
3.) Consider the grid of points to be plotted. How many point locations are in the scopeclock board memory array? How many times per second can this be accurately plotted on a 1Mhz analog scope without fading, considering the CRT's persistence and human persistence-of-vision? These are factors I feel that some software writers, who may treat a scope as a 'box', are unaware of. It's all very analog.
RANT OFF