Oscillation Sniffer

[Bonsai]

Hope this helps

 

[Elvee]

It is indeed a useful piece of kit.

I personally use slightly different versions of detectors, see below.

Here are some pics of the voltage-doubler version. It uses point-contact Ge diodes and is extremely sensitive.

I generally do not connect physically the probe to the DUT: it can upset the circuit, inject LF perturbations, or even cause oscillations because of the capacitances or ground loops.
I prefer to use a small sniffing antenna made of a centimeter or two of insulated wire.
I also use inductive sniffing loops, when mainly currents are involved (sometimes, oscillations generate rather low voltages on accessible nodes, and inductive sensing is preferable; it also pinpoints the radiating tracks more accurately).

 

[jan.didden]

What happened to the true and trusted RF detector scope probe??

Jan

 

[rsavas]

What happened to the true and trusted RF detector scope probe??

Scope probe method is okay for VHF range, not so good for UHF+, unless you have a GHz scope and a fet probe = very expensive.
The passive diode detector will go into the GHz range but has limited dynamic range!!
Now add a wide band sensitive rf amp, (BF998 or j310 in common gate config for example) to that passive diode detector and you really are getting much better for a very cheap solution.
We used to use a vhf/uhf rf pager at Motorola to detect rf leakage from rf signal generators. Damn those pagers were sensitive.

 

[Elvee]

Scope probe method is okay for VHF range, not so good for UHF+, unless you have a GHz scope and a fet probe = very expensive.

The RF detector probe mentioned by Jan is just that: a RF (diode) detector, very much like the first version of the circuits I showed. It was designed precisely to be able to detect tens or hundreds megahertz signals using a 2 or 4MHz oscilloscope.
However, for the more modern use of sniffing parasitic oscillations, this kind of old style probe is not ideal: the shapes described by Bonsai and myself are more adapted and flexible

 

[dchisholm]

There are several, commonly available, Schottky diodes intended for UHF detector applications. They should equal the venerable 1N34 (or 1N21) for sensitivity, even when you aren't looking for signals at 1 GHz and above. Designations include the BAT68, HSMS2850 and HSMS2860, and MBD101.

Dale

 

[SyncTronX]

There are several, commonly available, Schottky diodes intended for UHF detector applications. They should equal the venerable 1N34 (or 1N21) for sensitivity, even when you aren't looking for signals at 1 GHz and above. Designations include the BAT68, HSMS2850 and HSMS2860, and MBD101.

Dale, glad you posted that. I didn't want to order some diodes
for the OA81/OA91 That Bonsai posted in is sniffer.

But the good news is I have a supply of 1N34 that I don't use much.
So it looks like it should work. I'm going to be swapping some opamps
around and the sniffer looks like it will do the trick.

Thanks, y'all.

 

[Elvee]

Here is yet another version.

I also have variants with banana plugs, to use with a multimeter.

 

[peufeu]

it works the other way around, too.

For example, connect one end of coax to network analyzer output, nothing at the other end, just a bit of wire acting as an antenna, and walk it around the circuit without touching anything. Observe the output of the DUT (or any other interesting circuit node) while sweeping the frequency.

You expect coupling rising with frequency (capacitive), but if you get a medium-high Q boost like shown in the attachment, you got a problem somewhere. For example, the power supply shown below (an audiophile POS) actually amplifies line noise around 200-300k instead of rejecting it

Note you don't need a network analyzer. You can also inject a fast square wave and check for ringing at the output.