How to Build Your Own Oscilloscope Probes

Sure, you could buy oscilloscope probes off eBay for just about US$ 10. But where's the fun in that?

Besides, on eBay you might be gouged US$ 30 or more for shipping and "handling"...

Design

You will likely have problems if you directly connect an oscilloscope to a circuit under test using a coaxial cable with clips on the end. Oscilloscopes typically have input impedances on the order of 1MΩ. Meanwhile, 50-75Ω coax typically has a capacitance of 20-30 pF per foot. That means a very low impedance at radio frequencies, with two negative effects:
— Significant loading of the circuit under test
— Significant distortion of the measurement

Something like a 10:1 resistive divider probe circuit is commonly used to avoid these problems, here is a simple one:


             R1
            5 Mohm                 ----------------------
test <----/\/\/\/\----+-----------(o)                    )--- scope input
point                 |            +--------------------+     ~ 1 Mohm
                      |            |      50-75 ohm     |       10-20 pF
                      +--/\/\/\/\--+        cable       |
                        1.25 Mohm  |                    |
                           R2      |                    |
                                 -----                -----
                                  ---                  ---
                                   -                    -

       \__________________________/ \___________________/ \_____________/
                    |                         |                  |
                  Probe                     Cable           Oscilloscope

Why is this a 10X probe? R2 in parallel with the scope input impendance is:
1/(1/1.25 + 1/1) = 0.555555....
And so the ratio is:
(5 + 0.55555....)/0.55555..... = 10
The 1.25 MΩ resistor at R2 is a series combination of 1 MΩ and 250 kΩ resistors. Dealing with junkbox resistors (wait until you see the probe body!), the imprecision of your resistors and scope input impedance come into play. You could:
— Worry about the precision
— Call it "close enough"

Alternatively, here is a lower (but still reasonably high) impedance 1X probe:


             R1
           10 kohm           ----------------------
test <----/\/\/\/\----------(o)                    )--- scope input
point                        +--------------------+     ~ 1 Mohm
                             |      50-75 ohm     |       10-20 pF
                             |        cable       |
                             |                    |
                             |                    |
                           -----                -----
                            ---                  ---
                             -                    -

       \____________________/ \__________________/ \_____________/
                 |                     |                  |
               Probe                 Cable           Oscilloscope

For more on probe circuits, see these pages:
— http://www.sigcon.com/Pubs/straight/probes.htm
— http://www.repairfaq.org/sam/hvprobe.htm
— http://emcesd.com/1ghzprob.htm
— http://www.pomonaelectronics.com/index.php?i=a_scope_probe

Construction

Here's how I built my own oscilloscope probes. The images are thumbnails, click on one if you want to see a larger version.

I had some promotional pens with an intriguing design. The business end has a rubber-coated bulge making them pleasant to use as pens. So why wouldn't they make pleasant oscilloscope probes?
Here's another view.
Here is the complete bill of materials: The pen A 2-meter piece of coaxial test cable with a BNC connector on one end Epoxy adhesive One alligator clip Copper-plated nail — 0.75" (20mm) long, packed as "weather-stripping nail".
Start by disassembling the pen. Keep the pen main body and large body tip, the two large translucent pieces at left. Discard the remainder. Unless you have a use for those pieces...
Slip the upper barrel onto the cable, oriented so the free cable end is toward where the pen point used to be. Put another way, the cable running to the BNC connector exits from the end where the push button used to be. You will hold the pen as if you are writing when you use the probe. Strip back the outer jacket.
"Comb" the braid to open the braid without cutting or breaking the individual wires.
Twist the braid around the lead of a 1 MΩ resistor, and solder the lead into place. See the above about using a series combination of 1 MΩ and 250 kΩ for a more precise 10X ratio. I am using a single resistor in this version. This is R2 in the schematic above.
Trim the braid.
Connect a 5 MΩ resistor as R1 to the R2 resistor. Again, I am using just a single resistor as R2. Strip back the cable dielectric even with the far end of the R2 resistor. Wind the center conductor around that R1-R2 junction, and solder the junction into place. This is 75-Ω coax with a stranded center conductor — that makes this step easier.
Wind the far lead of the R1 resistor around the nail shaft near its head, and solder that junction.
Another view of things so far.
Mix the two-part epoxy and stuff it into the pen tip. The tricky part was getting the viscous epoxy into the tip. My solution was to "wind" as much epoxy as possible around the large toothpick I used to mix it, and let it drip into the pen tip. After several iterations of that, use the toothpick to work the epoxy into the tip. In this picture I have pushed the probe (nail) into and through the tip, and wiped the epoxy off the probe.
Slide the pen body into place, mark the cable jacket, and slide the pen body back up the cable. Cut the jacket off to allow a connection for a ground clip wire.
Slide the pen body back into place, and screw it into the pen tip. Solder a length of teflon-insulated wire onto the jacket.
Solder an alligator clip to the ground wire. Mix another small puddle of epoxy and work it into the upper end of the pen body, between the pen body and the cable.
Ready to use! Click here to see my page about repairing that Tektronix 2445A oscilloscope.