Guru
Joined: 09/08/2007
Location: AustraliaPosts: 4406
| Posted: 04:19pm 05 May 2017 |
Big fat primary is always going to be better. The more copper you can stuff through the hole the higher the power rating, and the greater the efficiency. Its a win win.
Skinny wire is fine for a first initial test to determine a suitable number of turns to get the output voltage into the desired range. Nothing wrong with joining up a few odd lengths of wire for testing.
Big round wire like welding cable or battery jumper leads are a simple answer, but it does leave a lot of large unusable voids between the turns.
Multiple skinny wires in parallel are always going to be the best way to get the maximum amount of copper into the available space.
Zero cost means recycling something, so whatever you can beg, borrow, or steal will be it. You could try your local friendly smiling scrap metal merchant, he probably has bins piled high with copper wire out the back, which he will be quite happy to overcharge you for. But its still going to be cheaper than buying new.
About salvaged grid tie chokes.
A while back I tested one of these, and it had two windings each of 42 turns, and it saturated just above thirty amps. I cannot now remember the exact figure, but can always do a retest.
Anyhow, that is 84 turns x 30 amps or 2,520 ampere turns. That will be a constant for that particular core with the particular air gap that it has.
So its pretty easy to work out where a particular number of turns on that specific core are going to saturate, without actually testing it. For example, ten turns would saturate at 252 amps if those original figures are correct.
Again, inductance is proportional to turns squared. So if we know that 84 turns gives us (maybe ?) 3,000 microhenries, we can work out the inductance for one single turn.
One turn would be 3,000 divided by 84 times 84 = 3000/7056 = 0.425 microhenries per turn.
With that figure, and knowing that inductance equals turns squared, ten turns would be 0.425 microhenries x 10 x 10 = 42.5 microhenries.
So its possible to rewind one of these grid tie chokes with far fewer turns, and make what used to run at a few hundred volts work at a few tens of volts at similar power levels.
This is all a pretty safe assumption. But if you need more, you could run more than one choke. Either two dual winding chokes in series, or one individual choke on each side.
It a case of playing with the numbers, working out how many turns can fit onto the core, and by trial and error coming up with something practical that has enough inductance and a suitably high saturation threshold to safely do the job.
I have a deep suspicion that the Chinese know little about all this, they simply directly copy something that appears to work, without really going through the actual engineering and testing.
Cheers, Tony.