I have a 20V, 2 Amp Max, 10Hz sine wave generator, I would like to be able to connect a transformer to step up and down the voltage.

How would I go about finding a transformer that I could connect to this and step up by a factor of 10, or down by a factor of ten?

What do I need to look out for using a transformer at such low frequency?

I have purchased a Triad TCT50-03E07K and tried to connect the generator to the secondary and on the primary I saw looked like the transformer was saturating, the current would spike like I have seen before with saturated inductors and transformers.

  • \$\begingroup\$ Could you use an amplifier instead? I am afraid the transformer will be extremely large and heavy, and maybe custom fabricated. Maybe an audio transformer from a tube amplifier? \$\endgroup\$
    – user57037
    Jun 2, 2016 at 0:18
  • 1
    \$\begingroup\$ To operate at such low frequency the transformer must have very high inductance which translates to huge core volume and copper volume.... I can't say how big, probably a cubic foot at least for 40 watts. \$\endgroup\$ Jun 2, 2016 at 1:02
  • \$\begingroup\$ To be honest, I know of no transformer that can pass frequencies below 10HZ, and that is out-of-band for even the best audio transformers. 10HZ is a valid value with some power penalty, but below 10HZ the core losses are extreme. \$\endgroup\$
    – user105652
    Jun 2, 2016 at 2:25
  • 1
    \$\begingroup\$ @Sparky256 Core losses increases with frequency. Lower frequency - lower core losses. See here:convertertechnology.co.uk/images/51.jpg The cross section area needed not to saturate it on the other hand increases with decreasing frequency. \$\endgroup\$
    – winny
    Jun 2, 2016 at 8:18

3 Answers 3


You can use a mains transformer at lower frequency, but you have to lower the input voltage to keep it below saturation.

At normal operating frequency the inductive reactance of the primary winding is sufficient to prevent the current from rising too high. At lower frequency the current has more time to rise so it goes higher and saturates the core, reducing inductance and causing the current to spike.

Reducing the voltage also reduces the rate of current rise, and so compensates for the lower reactance at lower frequency. Your 120V transformer should be OK with the primary winding powered by 20V at 10Hz, because that is 1/6th the frequency (10Hz vs 60Hz) but also 1/6th the voltage (20V vs 120V). Since it has a 10:1 ratio (120V in = 12V out) it will perform the other function that you want, 'step down by a factor of 10'.

However it will still only be able to handle the same current, which at 20V is 1/6th the power ie. 8.3VA instead of 50VA. To handle 2A you would need a mains transformer designed for 240VA (120V * 2A) which will be physically much larger. This size increase would also apply to a transformer that was designed to run at 10Hz (if such a thing existed).

To step up the voltage with a mains transformer the input winding must still be designed for 120V (at 60Hz). So what you want for stepping up the voltage by 10 times is a step up mains transformer with 120V primary and 1200V secondary, again rated at 240VA if you want it to handle 2A.

Edit to add:-

NOTE: Although a 240VA 60Hz transformer would be able to handle the power, it probably wouldn't do it efficiently. The power loss at 40W/10Hz would be similar to the loss at 240W/60Hz (perhaps a bit less due to lower magnetic losses at the lower frequency). If the transformer was 90% efficient at 240W/60Hz the loss would be 24W. That same loss at 40W/10Hz equates to just 40% efficiency. You might get the expected maximum amps, but at less than half the expected voltage.

To handle 40W/10Hz efficiently you would need an even larger transformer (or at least one with exceptionally low winding resistances for its size).

  • 2
    \$\begingroup\$ Sounds like a transformer from a junked microwave could be rewound for the task at hand.... \$\endgroup\$ Jun 2, 2016 at 2:25
  • \$\begingroup\$ You are correct in 1/6 of the frequency would mean 1/6 of the voltage to keep the voltage-time-area constant according to NAB=Vt, but the current will be unaffected by the frequency. OP needs to find a 6*20=120 Vrms, 2 A transformer or larger and he/she is good to go. \$\endgroup\$
    – winny
    Jun 2, 2016 at 6:41

20V at 2A is about 40W; this won't be a small transformer. You should be aware that all inductors have a volt-seconds limit (too many volts at low frequency saturates 'em), so your intention is to apply 2 volt-seconds (two volts, one tenth second) and that is a match to 120VAC at 1/60 second, i.e. 60 Hz line-voltage windings on common power transformers.

The second consideration will be the load current; if you actually deliver 40W to a load at 20A, the secondary winding has to have much lower resistance than 2V/20A. That means a 0.1 ohm secondary is too much secondary winding resistance. Similarly, 2A at 20V for the primary requires primary wiring resistance under 10 ohms.

At 40W, a 60 Hz stepdown x10 transformer would have limits of 120V/0.333A = 360 ohms (primary) and 12V/3.33A = 3.6 ohms(secondary); your lower frequency implies that you'd get the right I-squared-R scaling from a 120V, 60 Hz transformer that is rated for much more power (by a factor of 36). It'd be a 120V, 60 Hz, 1.4 kilowatt transformer.

So you need a factor of 6 in transformer power rating to keep from saturating, and a factor of 36 in transformer power rating to get just-as-efficient resistor power losses for your high currents.

There's a reason for switching power supplies; the transformers get cheaper for the high frequencies.


How about a vacuum tube audio output transformer, for example a Hammond 1650WA: http://www.hammondmfg.com/pdf/EDB1650WA.pdf This one is specified down to 30Hz at 280W, so should be just fine at 10Hz 40W. Your source could drive into the 16 Ohm "secondary", the full "primary" of 1900 Ohms is about 11:1 turns ratio (square root 1900 Ohms/square root 16 Ohms). But note there are a bunch of other primary taps that could be used if that ratio is too high.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.