Analog scope schematics anyone?

[Jebnor]

I'd like to see as many analog scope schematics as I can get my hands on.  So if you have any, or have links to some, could you post them please?

Why? I'm thinking of undertaking an educational project of building of building an analog scope. I know what you're thinking.

0) Hey good idea! Even if it doesn't work out, the schematics will be around, from which other can learn!
0.1) Why? Education
0.2) Why? Because I think it will be fun.
0.3) Why? What are we a 2 year old? move on...

1) What about the tube?  Once I get to the point where x-y-intensity coordinates of the trace are ready to plot, I expect to do a A-D conversion and send it to a graphics unit to display. Basically a digital tube, nothing more.

2) Why not build a DSO?  Because I want to learn more analog and, well, go big or go home.  Also, I have done data/signal processing in computers for years and I want something different.
2.1) What about cost, won't it just cost more than a decent DSO?  Maybe, see question 0.1. Education.

3) Okay...  How fast?  As fast as possible.  Being naïvely "realistic", I think a 20MHz scope would not be out of the question.  20yr old scopes are 20Mhz. I'm pretty sure I could make a reasonable approximation using the same chips, or possibly higher performance updates.
3.1) Number of Channels? >=2.
3.2) What features? Basic regular 2 channel 20MHz stuff. These details will get sorted as the research phase continues.

4) What about <insert item here>?  Don't know, you'll need to ask.
 
5) Pink is a colour right? well, no..  See minutephysics on YouTube.

[alm]

Tektronix and HP produced excellent service manuals with theory of operation and full schematics. Try looking and BAMA and download the manuals for some popular or interesting models. The Tektronix circuit concepts books are also an excellent read. Although old, they are probably the most detailed resource you're going to find on some of these topics.

1) What about the tube?  Once I get to the point where x-y-intensity coordinates of the trace are ready to plot, I expect to do a A-D conversion and send it to a graphics unit to display. Basically a digital tube, nothing more.
2) Why not build a DSO?  Because I want to learn more analog and, well, go big or go home.  Also, I have done data/signal processing in computers for years and I want something different.

This is probably going to be similar to early DSOs. Much of the magic in a good analog scope is in the CRT (which you're never going to replicate without just buying a NOS one). The vertical/horizontal amplifier is special because it requires substantial voltage swings to drive the plates. Without it, you're mostly left with an analog front-end and trigger circuit, something that's also common in DSOs. I guess you could digitize the time base ramp signal instead of a digital time base, but that would be the only difference I can think of. You still need a fast ADC for the vertical signal.

3) Okay...  How fast?  As fast as possible.  Being naïvely "realistic", I think a 20MHz scope would not be out of the question.  20yr old scopes are 20Mhz. I'm pretty sure I could make a reasonable approximation using the same chips, or possibly higher performance updates.

20 year old scopes could go up to 1 GHz, and that's not counting transient digitizers. That does not make it a trivial design exercise to build a 1 GHz scope though, even with modern parts. Don't count on using chips. Standard op-amps did not have the bandwidth back then, so they either used fully discrete parts or custom hybrids for the signal path. Of course you could redesign some of the circuits using modern op-amps.

Analog scopes also relied heavily on mechanical parts like impedance-matched attenuator switches. This is not something that is easy to replicate, unless you want to use FETs or relays as switches.

[c4757p]

1) What about the tube?  Once I get to the point where x-y-intensity coordinates of the trace are ready to plot, I expect to do a A-D conversion and send it to a graphics unit to display. Basically a digital tube, nothing more.

What? That's a DSO minus the storage... There are tons of old CRO tubes on eBay, just grab one. They're pretty easy to interface with if you don't mind a little HV.

I love the idea, though. I've toyed with the idea of doing a scope, too. I think I'll squirrel away an old CRT for a few decades until I retire, though...

[Jebnor]

1) What about the tube?  Once I get to the point where x-y-intensity coordinates of the trace are ready to plot, I expect to do a A-D conversion and send it to a graphics unit to display. Basically a digital tube, nothing more.

What? That's a DSO minus the storage...

My understanding is that in a DSO, the input level is converted to a level appropriate for the ADC, at which point the A-D conversion is done and a timestamp is added.  All the plotting is then done digitally. Is this correct?

I would like to plot using analog components.  The only thing the display unit would get would be (Trace A, x=0.284,y=0.824) or something to that effect.  The display unit would put the divisions on screen, and possibly the division ranges, mV/div us/div etc.

[c4757p]

Added complexity, really. You've basically got to have one more channel, entirely dedicated to capturing the X-coordinates, which you could just calculate from the timebase anyway. The vertical system would be almost entirely unchanged, they're basically the same thing in both types.