awesome video, keep on making videos about schematic breakdown , 🙏 greeting from your Indonesian friend 🇮🇩 🇺🇸

Do not think those diode bridges on the inputs are limiters, looks more like it is to deliberately introduce some deadband, so that the comparator will not generate multiple counts around the zero point of the waveform, or do spurious counts on too small a input signal. For small signals the output side diodes will be conducting, and the input will be biased, due to the diodes and the slightly asymmetric power supply voltages, to slightly below ground, and will be held there in a steady state, so making no output from the comparator. Input signal, after all the amplifier and gain control, has to exceed 0.6V, to get the one or the other input diode to conduct, shifting the top or bottom of the 2 series diode off of the value it is at steady state, and thus moving the comparator input away from the resting voltage, so it will change state rapidly, avoiding any spurious oscillations with slow input edges. The level control seems to bias the whole stage DC wise, so that you need more or less voltage to get past the threshold, as the level control is using a DC control voltage to basically adjust overall gain of the input stage, so that no matter the input signal you always have a very limited voltage applied to the comparator, but without needing to feed the input signal through variable controls, and allowing ATE equipment that is using this to get an input signal level to verify correct operation, as likely the external level control is also usable as input if selected as one. As to the reference oscillator, those likely are regular 10.7MHz IF stages, with some extra capacitance added to the primary side to tune them down, if they did not already have a wide enough tuning range to go down to 10MHz, which for these older larger designs is very possible, as they can easily tune 1MHz either side of the nominal. Double with the input side differential amplifier and diodes, then filter and amplify, apply AGC using the transistor acting as a as variable bias for the second amplifier, and do more filtering, so you get a good low distortion sine wave out. Then use the exact same comparator circuit again, to get the signal into digital form, so that any impurities in either one are going to affect both the same way. Gets rid of any subtle phase shift around the zero crossing in both conversions, as there now is a fixed minimum level, thus a well defined slew rate, where the comparator will be operating in.

So that is how a frequency doubler works, always meant to look that up sometime. Well one circuit anyhow, I'm sure there are many depending on frequency and other requirements.

I am lucky enough to have one of these. Peter at TRX Bench/Lab repaired his on his KZfaq channel. thanx.

Looks like a 1968 design, using tricks needed as fast logic gates were not available, then morphed with IEEE488 optional output.

Thank you for all the awesome videos! I just have a comment/question… do you have some kind of noise removal filter added on your audio track? Lately it seems like the intelligibly of the audio is worse than before. Perhaps you can tweak some settings so it is a bit easier to listen to. It’s just a minor thing, but it is sometimes difficult to hear what you are saying.

Copying channel A to B allows you to measure time interval events for the same line (time between 2 positive edges, for example). I have a 1994. Great machines.

watched 10 seconds, subscribed

Very nice, but I don’t understand why double the ref freq instead of generating the desired one - doubling would double any errors. Or is a stable 5Mz easier to generate and send than a 10Mz?

In the high frequency path between C79 and R113, there is a dog leg, then a circle, then a junction and the label "7,8". Is the dog leg a feature, or is it just drawn that way? What is the circle? Are these related to the label RLE1? Are they related to the nearby floating bit where C87 is between what looks like +14 volts and ground?

I wonder why they used single pole relays instead of switches.

I've heard that 5 MHz xtal references are ultimately more stable than 10 MHz, and thus the freq standard industry is moving from 10 MHz references to 5 MHz. (This from a guy who is involved with the Time-Nuts web mailer.)