I have seen planar magnetics transformers even with isolation up to 20 kV. However those devices seem to be designed for DC/DC conversion in 100 kHz range.

I am wondering whether low frequency laminated transformers can be replaced by a planar transformer?

In other words are there planar transformers which are suitable to convert 230VAC 50Hz to 20VAC (as an example)?


It would be extremely difficult because of the number of turns required for the primary (hundreds). If you had an expensive board with many, many layers then maybe.

To get a certain magnetizing current, the primary inductance required is proportional to the operating voltage and inversely proportional to the operating frequency. So for a frequency lower by a factor of 100kHz/50Hz, you need an inductance 2000 times higher, or about 45x as many turns (square root). So if a 10 turn primary would do at 100kHz, then a 450 turn primary is required at 50Hz.

Mostly the available cores are ferrite, so they would not be very appropriate for mains frequency either. Laminations of silicon steel are normally used at mains frequencies.

It may in fact be possible, but probably not economical.

  • \$\begingroup\$ Could you provide a little more information why that high number of turns is required (compared to the relative few for DC conversion applications)? And what would be the better core material for mains frequencies? \$\endgroup\$ – Pro Backup May 21 '15 at 19:36
  • 2
    \$\begingroup\$ At 50Hz you need much higher primary inductance, hence the high number of turns, and you probably can't find better than silicon iron laminations for the core. \$\endgroup\$ – Brian Drummond May 21 '15 at 19:43
  • 1
    \$\begingroup\$ Ha.+1 Brian beat my edit. \$\endgroup\$ – Spehro Pefhany May 21 '15 at 19:47

If the power requirements/expectations are low from the planar device it will work as an isolator but, as has been mentioned in Spehro's answer, the high inductance required to connect it directly to 230V AC, means the number of turns are unfeasible. You need high primary inductance to avoid core saturation.

Consider a standard mains transformer with a primary magnetization inductance of 10 henries has an impedance of 3142 ohms at 50Hz. To get this inductance it may have 1,000 turns ball-park. The primary mag current is 230V/3142 = 73 mA.

Ampere turns are therefore 73.

If the transformer only had 500 turns the inductance would be one-quarter (inductance in a coil like this is proportional to turns-squared) and the reactance would be 786 ohms. Now the current is 292 mA and, more importantly the ampere turns is 146.

Ampere turns create saturation and, it can be seen (hopefully) that fewer turns are going to rapidly lead to disaster in terms of performance. One or two folk may be murmering that "surely, the ampere-turns are far more when the transformer is under load and that the above is a pointless calculation" (or words to that effect).

No, saturation occurs at its maximum when a transformer is unloaded. Ampere turns due to load in the primary are cancelled by ampere turns in the secondary.

Anyway I'm going off at a bit of a tangent. What I'm trying to say is that you can use a planar transformer at 50Hz but connected to the AC via a capacitor. You can't get much power from it but if you can get enough turns on the primary to (say) work at 50V AC you can engineer a solution where the cap drops 180V.

There are of course load dependancies with this arrangement but, if the output is regulated then you can get a few watts from the device easily.

Who would do such a thing? I really don't know but it is possible and, maybe there is some application out there crying for this as a solution.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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