Build Your Own Transformer: EE Student's Guide

[mishimaBeef]

Hi, I was thinking of making my own power supply and I was wondering how low level I can go with it in terms of designing individual components and then putting them together to form the circuit.

So, I'm looking at like a 120V to 15V transformer then through a rectifier to get 15V DC?

I'm sure it's more involved than that but ideally that is all you need right? Perhaps some practical knowledge is being missed and if anyone can point that out great.

Now to the point, is it feasible to design and build my own transformer using like some iron and wiring the primary and secondary myself? Is there a way to do this safely (i.e. overdesign the insulation and accept low efficiency, fuse the primary).

I'm not looking to make a transformer that will be part of the final design of the power supply but I'm looking to test, analyse and explain the home brew design with experimentation.

Thanks,
EE Student

 

[KingNothing]

I'm sure it's more involved than that but ideally that is all you need right?

Not quite, you will get severe rippling without a capacitor to smooth it out.

So, I'm looking at like a 120V to 15V transformer then through a rectifier to get 15V DC?

Rectifiers have some loss - the simplest type, a full-bridge rectifier, will have a loss equal the forward voltage of two diodes, or about 1.2-1.4V for PN junction diodes. You will need to design around this: the output of your transformer should be around 16.4V.

Now to the point, is it feasible to design and build my own transformer using like some iron and wiring the primary and secondary myself? Is there a way to do this safely (i.e. overdesign the insulation and accept low efficiency, fuse the primary).

Yes, it is feasible, but be careful: use a fuse, use a GFI-protected outlet to test, and make sure you are using the 'ground' plug of the wall outlet to ground a metal casing around your circuitry. That way, if anything should go awry, it will do so via a blown fuse or circuit breaker, not a blown EE student!

Perhaps some practical knowledge is being missed and if anyone can point that out great.

Transformer turns ratios are not as cut-and-dry as a textbook makes you think. The "actual" turns ratio might differ from the ratio of voltages it gives you, and is slightly frequency-dependent. Build your design so that you can repeatedly test the ratio at 60Hz, and add turns as needed. You may have to do this 10 times to get it perfect!

Good luck.