coil winding for transformer/inductor

Thread Starter

jamjes

Joined May 10, 2010
35
Has anyone here wound their own components?

I'm experimenting by winding my own inductor (and eventually a transformer) by using a ferrite core and a reel of enameled wire.

Obviously winding by hand is pretty slow. How much does it matter if the coil isn't perfectly uniform and 'straight'? Like if the wires cross back and forth? Would it affect the inductance?

Any thoughts welcome
 

Bychon

Joined Mar 12, 2010
469
The wires don't HAVE to be all neat, but it helps.

I usta have a transformer guy, and he told me, "Engineers think transformers are made of iron and copper, but they have more pieces of masking tape and paper in them than all the iron and copper put together."

Take a hint from a pro. (Not me, the transformer guy.) Grow really thick fingernails and use lots of masking tape.
 

Thread Starter

jamjes

Joined May 10, 2010
35
Of course it will affect the inductance.
Well, I guess that is true :p but if the wires cross at say, 1 to 5 degrees, just how much difference will it make? If its less than 10% of my calculated value, then I'm probably okay with that for the purposes of experimenting. I'm just wondering if the result is severe.

The wires don't HAVE to be all neat, but it helps.

I usta have a transformer guy, and he told me, "Engineers think transformers are made of iron and copper, but they have more pieces of masking tape and paper in them than all the iron and copper put together."

Take a hint from a pro. (Not me, the transformer guy.) Grow really thick fingernails and use lots of masking tape.
Lol, I found that to be true when deconstructing a couple of transformers. Personally, I have found blobs of epoxy resin to work well, with tape on top. Does electrical insulation tape hinder flux?
 

Bychon

Joined Mar 12, 2010
469
No, tape and paper and rosin aren't magnetic. They don't bother the flux. Somebody will probably get on here next and tell you that if you put in so much paper that the windings are clear out to Tuesday, that will interfere, but we both know the paper and tape don't make the windings stick out all funny.
 

SgtWookie

Joined Jul 17, 2007
22,230
Using the proper tape (not "electrician's tape") or a bobbin will delay the onset of core saturation, so that's a desirable characteristic.

Have a look at Ronald Dekker's page:
http://www.dos4ever.com/flyback/flyback.html
It's mostly about DIY boost and flyback converters, but has lots of stuff on testing inductors.

Download/install the Mini Ring Core Calculator:
http://www.dl5swb.de/html/mini_ring_core_calculator.htm
It's a great (and free!) utility for ferrite and iron toroidal cores.

I made an inductor tester from a 74HC04 hex inverter, three 100nF caps and a couple of resistors. I use my o'scope to measure the PRT of the output square wave, and plug the results into a spreadsheet I made up that gives the inductance and AL value of the toroid.
 

SgtWookie

Joined Jul 17, 2007
22,230
It helps a great deal if you keep the turns wound on a toroid as evenly distributed as possible. The more evenly you have the turns spaced, the higher the Q of the coil will be. You can increase the inductance of a toroidal inductor somewhat by squeezing the windings together towards one side, but that really kills the Q.

One thing you really want to avoid is accidentally sticking your wire through a previous loop, which effectively makes a half-hitch knot. It messes things up quite a bit.

With low frequency stuff (say, 200kHz and under) the parasitic capacitance due to the closely wound magnet wire doesn't matter all that much. However, if you're doing HF RF stuff, the parasitic capacitance will have a rather large effect on the overall impedance and self-resonance of the inductor. Some folks use a "basket weave" for those types; you'll use lots more wire for the same inductance, but you'll have much lower parasitic capacitance.

Anyway, here's the schematic and board layout I used for my tester. For the LX1 and LX2 connectors, I used a pair of alligator clips screwed right to the board so as to add minimal inductance.

It works fairly well for what I need to do. Accuracy isn't great below around 10uH. I measured some 970nH (0.97uH) inductors that I'd tested on a $20k material analyzer, the rather crude method I used came up with around 2uH. Caveat Emptor.

Next time around, I'd use a 74VHCU04 IC and all SMT parts instead of a mixed bag of SMT and thru-hole..
 

Attachments

Last edited:
Top