Tektronix 475 - The Beats Keep Coming
A friend recently gifted me a Rigol spectrum analyzer and I've been having fun with it diagnosing and fixing problems with one of my antennas and looking for RF interference in the house, but that's for a future post. While that Rigol is nearly a decade old, it's far newer than my other test equipment, in some cases by more than 40 years. Much of my test equipment is from the era of bell bottom jeans and Starsky and Hutch, because A) I'm generally cheap, and B) I just don't have that much disposable income to dispose on a hobby. My experimenter board and oscilloscope are from the glorious 1970s, and my vacuum tube voltage meter is from the 1960s. Summers were longer, Spring was greener, young adults were strung out. Ah, the good old days.
I restored the non-functioning multi-meter a few years back, but the experimenters board and oscilloscope are living on with what they were born with.
That new shiny spectrum analyzer with its digital goodness and assumed accuracy got me to wondering if my old "Craigslist find" oscilloscope is still accurate enough for my amateur uses. I figured that the Tek 475 had probably drifted in accuracy, but when I was measuring the output of the function generator on my equally old Heathkit Electronic Design Experimenter Model ET-3100, the max frequency reported was nearly 65% higher than what the ET-3100 was supposed to be generate.
That Heathkit experimenter board's signal generator was never intended to be accurate but 65% is a huge deviation from the expectation. So I decided to measure the accuracy of my oscilloscope.
How to test?
Not having an actual signal generator I thought, well what can I trust to generate a known frequency that I can measure? The Tek 475 has a bandwidth of 200 MHz. So testing a signal near the top of that range should prove challenging for the old scope and give me an idea of accuracy on the time scale side of things because that will push it to the limit of its abilities. Regarding voltage measurements, I'd already compared it using a few stable voltage sources (batteries) against my multi-meter. They both measured exactly the same.
So, back to time-domain accuracy. The 2m simplex calling frequency is 146.520 MHz. That's pretty near the top of the scope's bandwidth. If you key a FM transmitter with no input (no modulation) it will generate a carrier at that frequency. There will be harmonics but they are well away from the primary frequency. To measure that on the oscilloscope I need to know the time it takes a wavelength at that frequency to complete. A 146.52 MHz signal has a period of 6.83 nanoseconds (ns). That sounds difficult to measure.
The scope's fastest native setting is 10 ns/div (labeled as .01µs)
A single cycle of this signal (6.83 ns) will occupy roughly 0.68 divisions on the CRT, so less than one division, which may make it difficult to assess. But if we engage the X10 Magnifier Switch we get an effective time division of 1 ns/div. With the magnifier on, one full wave cycle will now stretch across 6.83 divisions (the little squares), creating a waveform that's relatively easy to measure.
I keyed the HT into a dummy load at its lowest power setting next to the oscilloscopes' probe and saw...
Note the time division at 0.1µs and the X10 magnification is pressed
Ladies and gentlemen that is 6 full divisions and crossing midway past the 4th hatch mark of the next division (representing 8) giving us very close to 6.83. I don't have the start of the waveform exactly on the zero crossing so if that were a nudge to the left it would be crossing the 6.83 at the correct location.
I'd venture to say they don't make'em like they used to
Conclusions
Is an oscilloscope from the 70's as easy to use as a digital scope that has markers? No.
Does a digital scope make you think about the actual length of a waveform? No.
Were engineers smarter in the 60's and 70's? Who knows. They certainly had to work harder to get an answer to a question, but then they weren't being interrupted by messages and alerts on their phone every... hold on I got a message...
The Tektronix 475 was a professional piece of equipment back in its day. It likely was used in a lab that kept it calibrated until it became obsolete, probably sometime toward the end of the 80's. When I picked it up, it was pretty much as you see. The panel wasn't very dirty, the screen was clean and unmarked, there were no scratches on the case and there were still plastic caps over the rear connectors. It seems to have lived a pampered life.
But, for the past decade the poor thing has been abused by someone (me) who doesn't know how to properly use it. It serves as my silent electronics Elmer as I fumble around with it trying to confirm or deny my little hypotheses as I build equipment and alternatively let the smoke out of equipment (did I tell the story of how much kinetic energy a power transistor on a 1-watt transmitter can produce?)
So, if you are a nascent electronics experimenter looking to get some test equipment on the cheap, and you want to actually have to learn what your measuring, and you like to have a potential 50-70 year old fire hazard on your desk, keep an eye out for an old Tektronix scope from the good old days.
