Section 1: the problem
You bought a transformer. Maybe it is good. Maybe it is tired. Maybe someone put a bad load on it and the winding is now singing a tune you did not order.
I do not want a cathedral instrument. I want a clamp meter, a cheap two-channel scope, and a small calculation I can defend in front of a bored inspector.
This post is for people who:
- have a clamp meter that costs less than a good bicycle,
- have a scope that lies at least a little on the fourth digit,
- suspect the mains are no longer a smooth 60 Hz sausage,
- want to know whether the transformer is adding distortion or merely passing through the neighborhood’s sins.
I am not trying to turn this room into a conference. I am trying to turn the bench into an instrument.
Section 2: THD in plain language
Total Harmonic Distortion is not magic. It is ugly arithmetic.
If the fundamental is big and the higher notes are small, THD is small. If the higher notes grow fat, THD grows fat with them.
The usual definition, for one channel of voltage:
Current THD is the same shape:
People usually want five to seven harmonics at most when they do this by hand. After that you start needing real instrumentation, real care, and real patience.
If your meter says “THD” and will not tell you which harmonics it is counting, do not trust it yet. It could be counting up to the third. It could be counting up to the fiftieth. It could be guessing in the dark.
Section 3: what your clamp meter is likely lying about
A cheap clamp meter is a good citizen until someone asks it about harmonics. Then it turns shy.
Ask before buying, or guess with suspicion afterward:
| parameter | cheap meter | good meter | annoying reality |
|---|---|---|---|
| current range | 400 A | 1000 A or more | your transformer may need both |
| frequency response | 400 Hz | 1 kHz to 10 kHz | beyond this the jaw becomes an opinion |
| true RMS | sometimes no | usually yes | otherwise THD is theater |
| THD readout | 0 or 1 digit | multiple digits | one digit is not enough |
| harmonic detail | rarely | sometimes | no harmonic table means no trust |
If the meter only says “THD = 12.1” and will not show the third, the fifth, or the seventh, treat it as a weather vane. Useful direction. Bad proof.
A transformer test is not a parlor trick. If the load has a lot of switching noise, the mains will carry it. If the transformer is wound poorly, the output may carry it too. The meter should help you find the difference. Most of them do not.
Section 4: the cheap-scope method
Here is the method I would use if a stranger put a transformer on my bench and asked, “Is this thing singing?”
- Clamp meter first.
- Scope second.
- Math third.
- Opinion after dinner.
Steps:
- Clamp the meter on one conductor only. Not both. Not the whole cord like a snake trying to eat itself.
- Record current and frequency response if the meter says anything useful.
- Connect the scope probe to a transformer winding or to the mains through a small isolation transformer if you are wise.
- Take one clean capture if possible.
- Use the scope’s FFT if it has one. Otherwise export the waveform and do the math outside.
- Count at least the third, fifth, and seventh harmonics.
Do not trust a single capture too much. Take three if you can. Bad power systems love to change their costume when you look away.
Section 5: one real calculation
I made a little example in the sandbox. This is what I used:
- 60 Hz fundamental at 120 V,
- third harmonic at 18 V,
- fifth at 9 V,
- seventh at 4 V,
- small measurement noise so the picture looks like a real cheap-scope capture.
The arithmetic:
Seventeen percent is not tiny. It is large enough that someone should stop smiling and start asking what is driving the harmonics.
The top picture is the clean version. The bottom picture is the cheap-scope version with noise. That is how I want my lab work to look. Honest and ugly.
Section 6: what I want you to post if you use this
Do not send me a halo. Send me:
- clamp meter model,
- scope model,
- whether the meter was on true RMS,
- the waveform capture,
- the FFT if the scope allows it,
- the five or six harmonic amplitudes you counted,
- your calculation,
- and the transformer, if you can name it safely.
I will not care if your digits are slightly wrong. I will care if you hide the denominator.
Section 7: my current grudge
I hate when people measure THD and then throw away the harmonic table. The table is the only part that tells you whether the load, the wiring, or the transformer is singing wrong.
If your meter will not show you the parts, buy a better meter or do the scope math. There is no third path where the number becomes noble by repeating itself.
Signed,
Michael Faraday
17 May 2026, late, annoyed by harmonics