Oscilloscope probe cable extension

[npelov]

The title is not accurate, but that's the shortest version of it.
What I need is to make an extension board to tap off the signal going into the oscilloscope and make it available for another tool - like logic analyzer. You know that logic analyzers often give false results when the signal is bad. Viewing the signal on an oscilloscope will show if there are problems that are hidden for the digital circuits. Yes, there are MSOs on the market but you won't buy a whole new instrument because you need one feature.
Because I don't quite understand impedance matching (which probably does not apply in this case) and high frequencies I'm asking for advice how I should do that. Let's assume that it'll be used only on oscilloscope with high impedance input (not 50ohm).
1. What should be the recommended cable length? Is very short cable going to be a problem (because I know there are cases that short cables can harm the signal)?
2. Will any coaxial cable do the job or it must have specific properties?
3. Let's say it's for digital signal. What kind of buffer should I use. Is some 74xxx schmitt trigger enough (does it have high enough impedance).
4. advices on the pcb - the buffer will probably put some disturbance in the signal how to minimize that.
5. Usually it'll be used with digital signals up to 10MHz, but I want to ask is it hard to make it work fairly well for higher frequencies - let's say 50MHz (the frequency of the digital signal, not bandwidth).

[Marco]

A 1x probe is really already unusable at that upper frequency for the scope, you will have to use a 10x probe ... how to buffer it and how to connect it to your logic analyzer depends on what kind of logic analyzer you have.

[kg4arn]

Why not use a DIY Zo probe?  Low capacitance, high bandwidth, 1K (21:1) 5K(101:1)

http://www.sigcon.com/Pubs/straight/probes.htm

[Marco]

He probably already has a 10x probe for his scope, that's not really the issue ... he can just build a box with two BNCs to put in between the scope and a 10x probe with a buffer for the logic analyzer, it just depends on the logic analyzer what he does then.

If it's a cheap logic analyzer he can either use a 10x amplifier or a comparator as a buffer. If it's a logic analyzer which uses 10x probing I think he could use an unity gain buffer if he connected it directly without using a logic analyzer probe.

[pinkysbrein]

Hmm I guess he could just bodge the scope and logic analyzer probes together at the tips as well

[npelov]

@Macro you got it right. If someone doesn't have 10x  probe - there are cheap <$10 100MHz-200MHz probes in chinese e-shops.

I thought schmitt trigger would help with the noice, but ... probably a comparator will allow setting of threshold voltage.

@pinkysbrein of course you can do that. And you can put a multimeter too. But it wouldn't be that nice andclean. I always have troubles of too many wires stick out of my projects, so I'm trying to reduce them.
And probably I can just tap out the signal without buffer, but I don't want the scope to interfere the LA and the LA to interfere the scope. Also I want to have the advantage of high impedance probe.

@kg4arn Nice article. I'll check it out

[Marco]

I thought schmitt trigger would help with the noice, but ... probably a comparator will allow setting of threshold voltage.

A general logic schmitt trigger won't be able to deal with the (10x) attenuated levels, you'd first have to amplify.

You can accomplish the same thing with comparators. You have comparators which allow you to set two voltages for comparison (window comparators), with programmable hysteresis and you can of course use positive feedback but that's probably not a good idea at 50 MHz.

[vk6zgo]

What kind of Oscilloscope are you using?

Some analog 'scopes have an amplified version  of the input waveform present at the rear,so can be used as a buffer amp for other instruments,whilst at the  same time displaying the signal.

This may be a better way of doing what you want,rather than trying to "break out" the original signal at low level.