This project is kind of going to be like a long-term tweaking thing, I think. Either that, or I'll get so obsessed about it that I'll finish it in a few weeks. I don't know. It's going to be like how mechanical people work on their car. An oscilloscope is just a vehicle for looking at electronic signals, so it's just as important to this field as cars are to anyone else.
For years (probably 7 or more) I've had a Heathkit IO-103 single channel ~10MHz oscilloscope, donated by a generous friend. Not exactly top-of-the-line, but enough to see what's going on. A few years later, I recieved another donation, a Tektronix 475, which I've been using ever since (for good reason!). In the mean time, the Heath sat on the shelf. I recall it was in a state of unworking, but it may've just been out of adjustment. Bored with a cold, last night I got it out and started messing with it.
This is it. Apparently it was owned by a Marv Bergwall in Rock Island. I got it from someone else, who may've got it from Marv, or someone else. I didn't ask. In any case, here it sits on my bench, sides removed.
This is the right flank, which covers the trigger, sweep and deflection amplifier circuits on the bottom, power supply on top, and power transformer and fuse in the back.
And the left side. Here lies the CRT (in its shield), the other side of the power transformer (ditto), and the vertical amplifier and deflection circuit.
Here's a closeup of the horizontal deflection amplifier, which I was messing around with last night. It seems to be overcompensated, producing maybe 20% overshoot on a fast squarewave (fed from the front panel horizontal input jacks -- ewww, bananna!). You'll notice a lifted zener diode, in front of the nearest output transistor (Q214). That's ZD207, which seems to be a 6V diode, to keep the output amp from saturating I guess. But the thing is, it was acting within the on-screen range, causing compression of the sweep towards the edges. Without, the sweep still has a really long tail (the incoming waveform is a reasonably straight triangle edge, so it's in the amp), but at least it covers the full range.
Some rudimentary waveform tests. This is a square wave of about 1.2Vp-p (0.5V/div vertical) and 20 kHz (10us/div horizontal). Horizontal deflection seems to be about right, at least on the left side, counting 5 div = 50us for one cycle. Moreover, the middle rising edge is pretty close to the middle between the adjecent edges, so it's linear. But on the right side, you see it scrunched together, so sweep is slowing down there. The leading edges look okayish, but is noticably rounded for the last 20%, so HF gain is down by about 20% vs. midband.
This is at 10us/div and 200kHz. Essentially, a magnification of the leading edge rounding. It's still hard to see the main slope, so ultimate bandwidth at least is reasonable. 200kHz = 5us in 4 div., so actual time base is around 12.5us/div. I don't exactly know how this manages to happen, as the only difference from the previous setting is a precision film capacitor, which I would assume is correct. The 25% error implies an additional fixed capacitance of about 1.25nF, which is absurd. At least it's linear, in this range at least. I can calibrate one range, but not all of them. This must change.
0.1us/div, 2MHz. Now the rising edge can be seen well. 2MHz = 0.5us, so I should be seeing 5 div per cycle. Oddly, at this point, there's 4 div/cycle on the right side, and about 2.8div/half cycle on the left. It averages in the middle, but makes the apparent duty cycle wonky, which isn't the case. The rising/falling edge is about 0.2-0.3 div, which is around 20-30ns. Not too bad, and typical for such a circuit, using average transistors (the outputs aside; high voltage, high frequency transistors aren't that easy to come by). I'll have to see about compensating that gain rolloff though.
Most simple, I can tweak the existing circuit until it works nicely, and I'll have three 'scope channels at the Bench. But there's so much to change on the existing circuit (no, I don't give it much credit ;) that I might as well retool the whole thing and use it as a powered CRT with space for things inside. There are things useful, entertaining, amazing and novel that could be done. To list a few;