- On "You just [/glow]can buy decent feedthrough terminators, DC block adapters und attenuators":
Yes, correct. Technically it would work but practical you come up with a length of 10cm or longer if you connect DC-Block and Feedthrough-Termination or Attenuator together. I don't want to load this weight on my oscilloscope BNC input connector. Also this setup is only safe up to 1W RF and not 8W.
Most BNC 50 ohms pass throughs we have are 4cm, the BNC DC blocks we have are 5cm which makes 9cm, weighing around 35 grams each. If your scope's BNC can't hold a weight of 70 grams then it's a piece of crap and then you have a lot more problems than the lenght and weight of a BNC terminator and DC block
Also, BNC isn't the only format, you could use SMA componts which are even lighter and shorter.
And most SAs use N connectors which are even more massive and mechanically robust than BNC.
BTW, what's the weight of your friend's device?
- On "a solution in search of a problem."
One problem we had was to measure output ripple above 10 MHz from a DC/DC converter after the output filter avoiding direct coupling from converter into probe line. To avoid such coupling you need an impedance controlled wireing (e.g. 50 Ohm Coax) and termination close to your oscilloscope. Yes, we ended up in stacking DC-Block, Feedthrough-Termination in front of our oscilloscope and nearly break the input BNC connector.
Then I'm sorry but either your scope is crap or you're doing something very wrong. Also, you could have easily used SMA components if the mechanical load worried you.
I also wonder why you'd want to use a low impedance connection to measure ripple from a DC/DC converter.
The other problem was to check the output power of an Amplifier. We damaged a feedthrough termination by forgetting the DC-Block. We damaged a second feedthrough termination by putting too much power on it, not to mention RF-Transistors we had to replace as well.
Why the hell are you measuring amplifier output power through a scope connected with a low impedance line?
Standard passive probes have a bandwidth of 35 MHz (350 MHz on a well trimmed 1/10 divider). This is too low for fast transients.
Decent standard passive x10 probes are fine for up to 500Mhz, which is pretty much the limit for the class of scopes (Rigol DS1000/DS2000) you mentioned.
If you have a better scope with more BW then this will very likely also have an active probe interface for higher BW probes.