Keeping your world up and running.®

Extend your reach: Create a high-pass filter

By Chuck Newcombe

June 2007

Did you ever wish you had a more capable oscilloscope when a difficult problem presents itself? Well, I have. And I've learned over the years that sometimes you can make your simple 'scope or meter much smarter by the addition of a simple adapter. Here's an example.

Several years ago I was walking back to my desk from the break area when I was stopped by Bob, one of Fluke's 800 number responders. It seems he had a customer on the line who was having trouble trying to capture a switching transient, as the UPS he was testing switched from line to battery operation. Bob asked me if I had any idea how that might be accomplished using the dual channel ScopeMeter product that the customer was using.

My thoughts went back to my days in an aerospace lab, where I had access to a wide range of sophisticated oscilloscopes, including a dual channel one that had a high pass trigger filter to allow single-shot triggering by only high frequency glitches in the presence of a much larger lower frequency signal, such as the 60 Hz, 120 V signal the customer was trying to measure. Was there a way our customer could duplicate that capability?

I tried to explain what I had in mind to Bob, but he soon just handed me the phone and said, "Here. You explain it to the customer." So, there I was, explaining how to make the needed filter as an input accessory.

The customer was a technician in a product development lab, and had components such as resistors and capacitors readily available. I told him to get a 100 pF ceramic disk capacitor, rated at 250 V or more, and a 100 kOhm, ½ watt resistor, and then to connect them in series. I directed him to connect the free end of the capacitor to the line voltage output of the UPS, and the resistor to the neutral return point. Then, I had him connect the probe for the "A" channel of the scope to the junction of the two components and the common lead to neutral, reading the voltage across the resistor.

Finally, with the scope in the default free running mode, he said the 60 Hz signal was about 2.5 volts peak as the homemade high pass filter did its job. So, I instructed him to set the trigger point at about 3 volts and arm the single shot mode of the scope. Then, he unplugged the UPS, simulating a power failure, and noted that the scope trigger activated, displaying a snapshot of the transient that occurred.

Now it only remained to make use of the "B" channel of the scope. Its input was connected to the line output, in parallel with the filter input. The result was that the next time the glitch was captured on the "A" channel, its relative position on the full line voltage sine wave was clear. In this case, it occurred near the waveform negative zero crossing and had a peak magnitude of over 50 V.

A few days later, I decided to make an easy to use (and safe) version of the filter for my own use. I used a black rubber ac plug from my local big-box store and placed the components entirely inside. Now I can simply plug the adapter into any 120 V receptacle and connect the 'A' channel scope probe to the junction of the two components through the hole normally used for the power cord. Again, the common lead is connected to the adjacent neutral slot of the receptacle and the 'B' channel input to the unused 'hot' slot.

It was several months after I made the adapter that I was able to use it in a real application. The event that triggered the test first happened at 2:00 AM! That's right - I was up for a late night trip to the bathroom and noticed, as I turned out the fluorescent light when leaving to go back to bed, that the security light came on in our back yard.

I didn't investigate immediately. I waited until the next evening after the sun went down and set up my test connecting to the receptacle on the bathroom counter. I think it was on the third time I switched off the light that the outside light came on and, sure enough, my filtered single-shot triggered ScopeMeter input caught a transient- over 100V.

What was the cause? Well, the ballast in the fluorescent fixture is magnetic, and when the switch is turned off, the magnetic field collapses, acting like the spark coil on a car - a fast high voltage arc that flashes across the opening switch contacts.

How many of you have jury-rigged a test adapter from time to time in order to extend the capability of your 'scope or dmm to make a difficult measurement in a pinch? We'd like to hear about it.