3
\$\begingroup\$

I'm probing an RF (~75 MHz) oscillator. I'm trying to measure the voltage drop across a capacitor. (I'm afraid I may be pushing it's limits.) When I attach my x10 'scope probes it loads down the circuit some. (C probe = 16pF) Can I add a series capacitor (1-2 pF) to make another voltage divider and turn my probe into ~ x 100 (I'll have to calibrate it from another stiff source.)

Edit: So I tried it. Massive fail on the first attempt because of the inductance in the long ground lead of the probe... (long is ~10cm). I made a little ground clip. And soldered it on.

enter image description here

And that worked better.
Here's the 'scope shot with x10 probes top and bottom of the 2.2 pF cap that I have in series.

enter image description here

Channel 2 (in blue) is the "x100" side. I'm getting what looks like more bandwidth with the x100. This is not too surprising the x10 probe is an HP10071A Bandwidth = 150 MHz. But also the calibration factor is only about x30. 20 Vp-p vs. 7 Vp-p... Is that to be expected?

\$\endgroup\$
4
\$\begingroup\$

You can use a piece of paper or something similar, i.e. thin and non-conductive, like scotch tape. Place one piece over a ref.source and another one over the test point, then measure as usual and compare.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Careful using scotch tape, that stuff is an ESD generator. \$\endgroup\$ – Matt Young Feb 11 '15 at 21:38
  • \$\begingroup\$ I'm not following you? Can you explain a bit more? \$\endgroup\$ – George Herold Feb 11 '15 at 21:38
  • \$\begingroup\$ @George Herold: A piece of insulation material placed over the conductor acts as a capacitor in series with the probe tip placed nearby, as close to the conductor of interest as possible but not too close (don't poke through it). The capacitance of this makeshift cap depends on the thickness of the insulator of choice. The conformal coating or even oxide layer covering the surface of the conductor will act the same way but probing will require finesse and calibration will be harder. \$\endgroup\$ – Oleg Mazurov Feb 11 '15 at 21:52
  • \$\begingroup\$ @Matt Young - Anything non-conductive is an ESD generator. \$\endgroup\$ – Oleg Mazurov Feb 11 '15 at 21:54
  • \$\begingroup\$ @OlegMazurov got it. I've, wrapped a insulated wire around a lead, in the past. I was thinking if I had a "known" capacitance. I'll give the wire trick a try tomorrow. \$\endgroup\$ – George Herold Feb 12 '15 at 1:18
2
\$\begingroup\$

A series 2pF capacitor will make your probing setup frequency dependent, and subject to the poor tolerances of capacitors. You are unlikely to get an accurate representation of the actual waveform. Can't recommend this method.

The correct tool for this job is an active oscilloscope probe, or a low capacitance passive probe.

Your 16pF probe adds a 132ohm load at 75MHz, which of course is a significant load.

There is a poor mans alternative: A transmission line probe.

  1. Set your scope to 50ohm, 10X (or 100X).
  2. Connect a 50ohm coaxial cable to your scope input, and solder a ~450ohm (or ~4950ohm) surface mounted resistor in series with the coaxial center conductor and the DUT.
  3. Solder the coaxial shield to DUT GND. Keep the center conductor and ground attachment extremely short. The coaxial cable length does not matter much.

Now you have a 450+50=500ohm 10X probing setup (or a 1950+50=5000ohm 100X probing setup) Capacitive loading is equal to the capacitance of your series resistor, which is extremely low.

| improve this answer | |
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.