#1087 Do audio transformers add distortion? reprise

Episode 1087

[UltraSteaKME]

Nice! Now since we are talking about distortion, I would have seen the magnetic saturation effect when inputting more juice. This is what we typically refer as transformer distortion. Here we are just witnessing the effect of bad vs good matching impedances (and thanks for that!), which is not distortion but ringing.

[RexxSchneider]

Assuming the same wire is being used on both windings, you get the ratio between windings equal to 180:120 = 3:2. That means the resistance you place on the secondary is seen from the primary as (3/2)² = 9/4 times larger or smaller depending on which way round the transformer is connected. In one case, you loaded the secondary with 6K8 and in the other case you needed a load of 2K2. Assuming you're only able to estimate the matching impedance to around ±20% or so, that's perfectly consistent with required loading of somewhere around 4K on the primary (4k x 9/4 = 9K and 4K x 4/9 = 1.8K).

[russellhltn]

Actually, I'd expect a transformer to roll off the bottom end. As you get closer to DC, the rate of change falls off and it won't transform. Higher frequencies do carry over. You've demonstrated that at the higher end, the choice is between ringing and roll-off.

[borisj]

Distortion is caused by any non linearity. Transformer cores have two sources of non linearity. Hysteresis and core saturation. BH curves are non linear at the beginning of magnetisation and at the point approaching saturation. If you operate the transformer only between these in the linear section your signals will not suffer non linear changes, eg. Distortion.

[boonedockjourneyman]

You make my day better. Your genuine nature is so damn rare on YouTube.

Trying to hunt down ever bit of ring depends on how much noise vs signal attenuation matters. It’s engineers not science. People are flip-flopping around to get whatever answer they think they remember from school, reddit, whatever. Forget it. Your answers are as true for you today as they are everyday. You are the real deal. ••• and you don’t have that many years on me so •••

[byronwatkins]

Transformers also raise the lower frequency cutoff; they don't work at DC as an extreme limit. Also, the driver output impedance comes into play since tau=RC or tau=L/R. (Most precisely, superconductive windings do allow infinite time constant and DC response, but these are not superconductors.)

[richbecke]

To my knowledge, distortion of an audio signal may be audible at much lower levels than what you can visually discern by looking at a time domain scope display. It would be really interesting if you could repeat the tests and compare fft plots (ie. in the frequency domain) of the input and output signals. An 8-bit scope does not have the resolution to detect harmonics 60dB or more below the fundamental frequency. If you have a PC and a reasonably good audio interface, there's an excellent program called audiotester that lets you perform such measurements.

[ChiefMade]

Great reprise. Since you are a radio guy, have you ever used a transformer in a studio mic line box like the w2hy isolation boxes. An isolation box build would make a great vid. May do one myself, but just curious. Thanks for the video

[supercompooper]

reluctance is Ampere-turns per Weber 🤠👍

[montekay]

Check non linear distortion, your Rigol scope should have an FFT function.

[KeenanCrow]

Great video. I’ve always had this question as well.

[imgeekyou]

Great video. Maybe the signal integrity at higher frequencies is not so great because you need to do impedance matching on the primary side of the transformer as well.

[patwhens]

Always enjoy the vids! Add some trigger holdoff to keep the scope from triggering on the noise.

[thomasmaughan]

"Do audio transformers add distortion?"
No
(Betteridge's Law of Headlines)

[mortenhattesen]

Measuring/categorizing transformers using DC resistance isn't really the best way, as wire gauge comes into play. In this case, with a simple audio transformer it is ok, though.
But measuring the inductance is a much better way to easily categorize the primary/secondary winding ratios. If you have an LCR meter, it wouldn't be any more difficult than using the multimeter to measure the DC resistance.

[deanhedin]

Repeat test with air core audio isolation transformer

[InssiAjaton]

Are we talking about output transformers or input (like microphone) transformers? On most output transformers you take feedback to the amplifier from the secondary of the transformer. Thereby you attenuate any distortion by the amount of the feedback, maybe 20 dB which equals 1/10 in voltage. You generally never have that available on microphone transformers.
As to the measurements - it would have been much better to use two channels of the scope, one on primary and the other on secondary to demonstrate the potential differences caused by the transformer.

[vizkoze]

would adding a capacitor in paralell to the output resistor do anything positive to the ringing?

[blackarrow]

A question: If the transformer core would reach saturation, would the big chunk of steel then matter?

[mrenrollment]

Optionally if you want to extend this series - you could demo max power transfer matching.
I assume the generator is 50ohms. How to load the transformer to get the max power transfer, ...?