Dick Cappels' project pages http://www.projects.cappels.org

An adapter to allow low capacitance probing of high frequency circuits.

(Above) BEFORE AND AFTER. On the left, is the observed response to a 1 MHz to 20 MHz sweep of an amplifier when the output of the amplifier was probed directly with a Tektronix P2220 voltage probe. On the right, is the amplifier response observed when the low capacitance adapter was used.

Overview

My boss, Dave, said "Just hold the probe close to the cathode lead." Dave had worked at Tektronix for many years, and his ability to make difficult measurements was second only to his abilities to manage a company and design clever circuits. Dave's point was that, even though we didn't have a an FET probe, we could measure wide band signals at a fairly high impedance point in the circuit by capacitively coupling though the air to the wire carrying the signal we were interested in. The capacitance between the wire and the probe would form a capacitive voltage divider with the probe's input capacitance, and at frequencies significantly higher than the low frequency roll off created by the probe's input capacitance and the probe's resistance, the wave form would be faithfully reproduced, although it would have a reduced amplitude.

While working on a wide band amplifiers, I would often clip a small, 1 or 2 pf capacitor to my probe tip and use it to view the wave form at a high impedance point, such as the collector or drain of a transistor that is driving a resistive load. See the photographs above.  I realized that it would be more straight forward to just make a calibrated 10:1 capactive attenuator, so that I could then make calibrated measurements. This project is the old trick, made more convenient.

The circuit give here, presents about a 2 pf load to the circuit, and capacitively loads the circuit measurably less than the X100 voltage probe (A Stack CP-241) I often use for this sort of thing.

The capacitors, in conjunction with the probe's input capacitance, forms a capacitive voltage divider.

To make a X10 capacitive attenuator, simply select a variable capacitor, C2 that has an adjustment range that includes the capacitance, that when added to Cin, the input capacitance of the scope probe you are going to use with it, is nine times that of the coupling capacitor. In this circuit, the coupling capacitor, C1.  is 2 pf and the scope probe's input capacitance is about 16 pf, so C2 is

C2 = (9X C1) - Cin
C2 =( 9 X 2 pf ) -16 pf
C=  2pf

This set of values puts 90% of the voltage across C1 and 10% across C2 and the scope probe.  This also works for higher attenuations, but since I am working with signals of a few volts or less, X10 is about as high as I need to go.

Construction and Calibration

I built this on a small piece of pre-punched fiberglass board with a plated metalization pattern. The leads are all short to minimize stray reactances. A couple cm of bare #30 tinned copper wire serves as the probe. A small hook at the end lets me "hang" the probe on component leads. I also occasionally solder the tip of the wire to the circuit under test.

A short length of hookup wire makes the ground connection from the adapter to an alligator clip which makes the ground contact for the circuit under test. The shorter the better. For frequencies above 50 MHz, it would be much better to just solder a short piece of wire, about 1 cm long, from the adapter's ground to the circuit under test.

Two pieces of nickel plated paper clip are soldered to the board to serve as connection points for the scope probe and the probe's ground clip. A connector that accepts the probe body without the probe tip would be much better, and as I recall, Tektronix used to sell such connectors. In my case, I don't have anything suitable, so I used the nickel plated paper clip. Solder flux, in addition to that in the flux core solder, was applied to the paper clip while soldering. I think it helped a lot.

(Above) The component side of the board. Keep the layout tight.

(Above) The wiring side of the board. The blue knot at the bottom
of the board is wire wrapping wire I used to tie the gray ground
lead to the board as a strain relief.

The low frequency -3db point of this adapter is about 800 Hz (20 pf and 10 Meg Ohms), so to calibrate it, I connected the input to a signal generator set generate a sine wave at a much higher frequency, around 200 kHz, with 1.0 volts peak-to-peak, and adjusted C2 to obtain 100 millivolts peak-to-peak on the scope. I was not surprised to notice that capacitance from the scope probe tip to the ground lead and the circuit under test affected the calibration. To minimize this effect, physically arrange the adapter and scope probe to minimize stray capacitance to ground and the circuit under test. This is where that Tektronix connector for the probe would have come in handy.

HOME (More Projects including test and measurement projects)